[go: up one dir, main page]
Next Issue
Volume 8, August
Previous Issue
Volume 8, June
 
 
Search for Articles:
Title / Keyword
Author / Affiliation / Email
Journal
Article Type
 
 
Advanced Search
 
Section
Special Issue
Volume
Issue
Number
Page
 
6.9
2.8
Journals
IJGI
Volume 8
Issue 7
share announcement

ISPRS Int. J. Geo-Inf., Volume 8, Issue 7 (July 2019) – 26 articles

Cover Story (view full-size image): Thanks to the availability of geospatial data, 3D environment reconstruction has been used in the visualization of urban environments. OpenStreetMap offers a huge potential by providing a flexible, crowdsourced alternative to use for such purposes. In this work, City Maker, a 3D environment reconstruction tool, was designed to use OpenStreetMap as the map data source to generate digital city models. By providing additional parameters such as roughness coefficients and storm drains, the digital city model was made simulation-ready. By arranging the layers in an appropriate format, the city models were loaded and visualized efficiently. The model was tested with a hypothetical flooding scenario to demonstrate its applicability. It is also noted that given data availability, OpenStreetMap can be used to generate digital city models. View this paper.
  • Issues are regarded as officially published after their release is announced to the table of contents alert mailing list.
  • You may sign up for e-mail alerts to receive table of contents of newly released issues.
  • PDF is the official format for papers published in both, html and pdf forms. To view the papers in pdf format, click on the "PDF Full-text" link, and use the free Adobe Reader to open them.
Order results
Result details
Show export options expand_more Show export options expand_less
Select all
Export citation of selected articles as:
30 pages, 11880 KiB  
Article
Efficiency of Extreme Gradient Boosting for Imbalanced Land Cover Classification Using an Extended Margin and Disagreement Performance
by Fei Sun, Run Wang, Bo Wan, Yanjun Su, Qinghua Guo, Youxin Huang and Xincai Wu
ISPRS Int. J. Geo-Inf. 2019, 8(7), 315; https://doi.org/10.3390/ijgi8070315 - 23 Jul 2019
Cited by 9 | Viewed by 3615
Abstract
Imbalanced learning is a methodological challenge in remote sensing communities, especially in complex areas where the spectral similarity exists between land covers. Obtaining high-confidence classification results for imbalanced class issues is highly important in practice. In this paper, extreme gradient boosting (XGB), a [...] Read more.
Imbalanced learning is a methodological challenge in remote sensing communities, especially in complex areas where the spectral similarity exists between land covers. Obtaining high-confidence classification results for imbalanced class issues is highly important in practice. In this paper, extreme gradient boosting (XGB), a novel tree-based ensemble system, is employed to classify the land cover types in Very-high resolution (VHR) images with imbalanced training data. We introduce an extended margin criterion and disagreement performance to evaluate the efficiency of XGB in imbalanced learning situations and examine the effect of minority class spectral separability on model performance. The results suggest that the uncertainty of XGB associated with correct classification is stable. The average probability-based margin of correct classification provided by XGB is 0.82, which is about 46.30% higher than that by random forest (RF) method (0.56). Moreover, the performance uncertainty of XGB is insensitive to spectral separability after the sample imbalance reached a certain level (minority:majority > 10:100). The impact of sample imbalance on the minority class is also related to its spectral separability, and XGB performs better than RF in terms of user accuracy for the minority class with imperfect separability. The disagreement components of XGB are better and more stable than RF with imbalanced samples, especially for complex areas with more types. In addition, appropriate sample imbalance helps to improve the trade-off between the recognition accuracy of XGB and the sample cost. According to our analysis, this margin-based uncertainty assessment and disagreement performance can help users identify the confidence level and error component in similar classification performance (overall, producer, and user accuracies). Full article
►▼ Show Figures

Figure 1

Figure 1
<p>Eight study areas and the main existing land cover types (all classes in Beihai are in true colour mo de and classes in Hobart are in pseudo mode(R/G/B:4/3/2), expect 3 classes (road, highlight objects and shade). Dataset 1 of aerial images of Beihai is shown in a, and b is for dataset 2 of Geo-Eye 1 at Hobart.</p> Full article ">Figure 2
<p>Spectrum histogram of land cover types in Beihai (<b>a</b>): R band and Hobart (<b>b</b>): R band.</p> Full article ">Figure 3
<p>Overview of the workflow for extreme gradient boosting (XGB) with imbalanced data.</p> Full article ">Figure 4
<p>Overall accuracy curves of XGB (black) and random forest (RF) (red) for 8 areas across different minority proportions (minority: majority from 1:100 to 100:100) of the training data set. The results of different areas are at: (<b>a</b>) area 1, (<b>b</b>) area 2, (<b>c</b>) area 3, (<b>d</b>) area 4, (<b>e</b>) area 5, (<b>f</b>) area 6, (<b>g</b>) area 7, (<b>h</b>) area 8.</p> Full article ">Figure 5
<p>Stacked bars of overall disagreement performance of XGB (grey) and RF (red) for all 8 area across different sample imbalance. The results of different areas are at: (<b>a</b>) area 1, (<b>b</b>) area 2, (<b>c</b>) area 3, (<b>d</b>) area 4, (<b>e</b>) area 5, (<b>f</b>) area 6, (<b>g</b>) area 7, (<b>h</b>) area 8.</p> Full article ">Figure 6
<p>Reference map (<b>a</b>) and Classification maps of XGB using the training sets with the ratio of minority: majority at: (<b>b</b>) 1:100, (<b>c</b>) 10:100, (<b>d</b>) 20:100, (<b>e</b>) 30:100, (<b>f</b>) 40:100, (<b>g</b>) 50:100, (<b>h</b>) 60:100, (<b>i</b>) 70:100, (<b>j</b>) 80:100, (<b>k</b>) 90:100, (<b>l</b>) 100:100, for area 6.</p> Full article ">Figure 7
<p>Curves of producer and user accuracies for the minority (<b>a</b>: house) and one (<b>b</b>: soil) of the 5 majority classes for area 6 across different sample imbalance.</p> Full article ">Figure 8
<p>The <span class="html-italic">F</span><sub>1</sub> score curves of XGB and RF for all areas across different minority proportions (minority: majority from 1:100 to 100:100). The results of different areas are at: (<b>a</b>) area 1, (<b>b</b>) area 2, (<b>c</b>) area 3, (<b>d</b>) area 4, (<b>e</b>) area 5, (<b>f</b>) area 6, (<b>g</b>) area 7, (<b>h</b>) area 8.</p> Full article ">Figure 9
<p>Stacked bars of disagreement performance for the minority (<b>a</b>: house) and one (<b>b</b>: soil) of the 5 majority classes for area 6 across different sample imbalance.</p> Full article ">Figure 10
<p>Disagreement performance of the minority classes using XGB (grey) and RF (red) for all areas. The results of different areas are at: (<b>a</b>) area 1, (<b>b</b>) area 2, (<b>c</b>) area 3, (<b>d</b>) area 4, (<b>e</b>) area 5, (<b>f</b>) area 6, (<b>g</b>) area 7, (<b>h</b>) area 8.</p> Full article ">Figure 11
<p>The probability-weighted margin (PWM) cumulative frequency distribution curves of area 6 grouped with correctly and misclassified instance of XGB using the training sets with the ratio of minority: majority at: 1:100 (dark), 10:100 (orange), 50:100 (green), 100:100 (blue).</p> Full article ">Figure 12
<p>The boxplots for mean margin of XGB and RF using 10 independent trails across different minority proportions (minority: majority from 1:100 to 100: 100). The results of different areas are at: (<b>a</b>) area 1, (<b>b</b>) area 2, (<b>c</b>) area 3, (<b>d</b>) area 4, (<b>e</b>) area 5, (<b>f</b>) area 6, (<b>g</b>) area 7, (<b>h</b>) area 8.</p> Full article ">Figure 13
<p>The boxplots of margin entropy of XGB and RF using 10 independent trails across different minority proportions (minority: majority from 1:100 to 100:100). The auxiliary line (y = 3.32) represents the theoretical maximum of the margin entropy with a 10-bins distribution. The results of different areas are at: (<b>a</b>) area 1, (<b>b</b>) area 2, (<b>c</b>) area 3, (<b>d</b>) area 4, (<b>e</b>) area 5, (<b>f</b>) area 6, (<b>g</b>) area 7, (<b>h</b>) area 8.</p> Full article ">Figure 14
<p>Overall accuracy of XGB and RF for the experiments when tree class and water class in area 8 are assigned to be the minority class in the training sets, respectively, across different minority proportions (minority: majority from 1:100 to 100:100).</p> Full article ">Figure 15
<p>Overall disagreement performance of XGB and RF for area 8 when the minority class is of different spectral separability (<b>a</b>): water; (<b>b</b>): tree.</p> Full article ">Figure 16
<p>Minority class classification maps of XGB with different sample imbalance and spectral separability. The first row represents reference map (<b>a</b>) and the results when water class is the minority with different minority proportions (minority: majority) at: (<b>b</b>) 10:100, (<b>c</b>) 50:100, (<b>d</b>) 100:100. The second row represents the corresponding data (<b>e</b>: reference map for tree class) and results at: (<b>f</b>) 10:100, (<b>g</b>) 50:100, (<b>h</b>) 100:100 when tree class is the minority. Colours refer to different classes (water: blue, tree: green).</p> Full article ">Figure 17
<p>The <span class="html-italic">F</span><sub>1</sub> score curves for the minority class of XGB and RF with different spectral separability of the minority class across different minority proportions (minority: majority from 1:100 to 100:100).</p> Full article ">Figure 18
<p>Curves of producer and user accuracies for area 8 when the minority class is of different spectral separability ((<b>a</b>): water; (<b>b</b>): tree) across different sample imbalance.</p> Full article ">Figure 19
<p>Disagreement performance of specific class using XGB and RF for area 8 when the minority class is of different spectral separability (<b>a</b>): water; (<b>b</b>): tree.</p> Full article ">Figure 20
<p>The PWM cumulative frequency distribution curves grouped with correctly and misclassified instance of XGB when tree (green) and water (blue) in area 8 are assigned to be the minority class in the training sets, respectively. The results of different sample imbalance are (minority: majority) at: (<b>a</b>) 1:100, (<b>b</b>) 10:100, (<b>c</b>) 50:100, (<b>d</b>) 100:100.</p> Full article ">Figure A1
<p>Reference map (<b>a</b>) and Classification maps of XGB using the training sets with the ratio of minority: majority at: (<b>b</b>) 1:100, (<b>c</b>) 10:100, (<b>d</b>) 20:100, (<b>e</b>) 30:100, (<b>f</b>) 40:100, (<b>g</b>) 50:100, (<b>h</b>) 60:100, (<b>i</b>) 70:100, (<b>j</b>) 80:100, (<b>k</b>) 90:100, (<b>l</b>) 100:100, for area 8.</p> Full article ">
16 pages, 2405 KiB  
Article
Regionalization Analysis and Mapping for the Source and Sink of Tourist Flows
by Qiushi Gu, Haiping Zhang, Min Chen and Chongcheng Chen
ISPRS Int. J. Geo-Inf. 2019, 8(7), 314; https://doi.org/10.3390/ijgi8070314 - 23 Jul 2019
Cited by 12 | Viewed by 5527 | Correction
Abstract
At present, population mobility for the purpose of tourism has become a popular phenomenon. As it becomes easier to capture big data on the tourist digital footprint, it is possible to analyze the respective regional features and driving forces for both tourism sources [...] Read more.
At present, population mobility for the purpose of tourism has become a popular phenomenon. As it becomes easier to capture big data on the tourist digital footprint, it is possible to analyze the respective regional features and driving forces for both tourism sources and destination regions at a macro level. Based on the data of tourist flows to Nanjing on five short-period national holidays in China, this study first calculated the travel rate of tourist source regions (315 cities) and the geographical concentration index of the visited attractions (51 scenic spots). Then, the spatial autocorrelation metrics index was used to analyze the global autocorrelation of the travel rates of tourist source regions and the geographical concentration index of the tourist destinations on five short-term national holidays. Finally, a heuristic unsupervised machine-learning method was used to analyze and map tourist sources and visited attractions by adopting the travel rate and the geographical concentration index accordingly as regionalized variables. The results indicate that both source and sink regions expressed distinctive regional differentiation patterns in the corresponding regional variables. This study method provides a practical tool for analyzing regionalization of big data in tourist flows, and it can also be applied to other origin-destination (OD) studies. Full article
(This article belongs to the Special Issue Smart Cartography for Big Data Solutions)
►▼ Show Figures

Figure 1

Figure 1
<p>Study area (all cities of mainland China and all scenic spots in Nanjing).</p> Full article ">Figure 2
<p>Regionalization results of tourist flow intensity. The whole of China is divided into 15 regions, and each division is coded in Roman numerals.</p> Full article ">Figure 3
<p>Side-by-side boxplots for regionalization results of cities in China.</p> Full article ">Figure 4
<p>Regionalization result of the sink. (Note: The whole of Nanjing is divided into six regions, and each region is coded in Roman numerals.).</p> Full article ">Figure 5
<p>Mean line boxplots for regionalization results of scenic spots in Nanjing.</p> Full article ">
16 pages, 8419 KiB  
Article
An Agent-based Model Simulation of Human Mobility Based on Mobile Phone Data: How Commuting Relates to Congestion
by Hao Wu, Lingbo Liu, Yang Yu, Zhenghong Peng, Hongzan Jiao and Qiang Niu
ISPRS Int. J. Geo-Inf. 2019, 8(7), 313; https://doi.org/10.3390/ijgi8070313 - 23 Jul 2019
Cited by 28 | Viewed by 5084
Abstract
The commute of residents in a big city often brings tidal traffic pressure or congestions. Understanding the causes behind this phenomenon is of great significance for urban space optimization. Various spatial big data make the fine description of urban residents’ travel behaviors possible, [...] Read more.
The commute of residents in a big city often brings tidal traffic pressure or congestions. Understanding the causes behind this phenomenon is of great significance for urban space optimization. Various spatial big data make the fine description of urban residents’ travel behaviors possible, and bring new approaches to related studies. The present study focuses on two aspects: one is to obtain relatively accurate features of commuting behaviors by using mobile phone data, and the other is to simulate commuting behaviors of residents through the agent-based model and inducing backward the causes of congestion. Taking the Baishazhou area of Wuhan, a local area of a mega city in China, as a case study, we simulated the travel behaviors of commuters: the spatial context of the model is set up using the existing urban road network and by dividing the area into space units. Then, using the mobile phone call detail records of a month, statistics of residents’ travel during the four time slots in working day mornings are acquired and then used to generate the Origin-Destination matrix of travels at different time slots, and the data are imported into the model for simulation. Under the preset rules of congestion, the agent-based model can effectively simulate the traffic conditions of each traffic intersection, and can induce backward the causes of traffic congestion using the simulation results and the Origin-Destination matrix. Finally, the model is used for the evaluation of road network optimization, which shows evident effects of the optimizing measures adopted in relieving congestion, and thus also proves the value of this method in urban studies. Full article
(This article belongs to the Special Issue Spatial Big Data, BIM and advanced GIS for Smart Transformation: City, Infrastructure and Construction)
►▼ Show Figures

Figure 1

Figure 1
<p>Procedures of assigning a user to a spatial unit.</p> Full article ">Figure 2
<p>Diagram of rules for resident Agent behavior.</p> Full article ">Figure 3
<p>Road network and division of urban spatial units. (<b>a</b>) The original blocks and road network; (<b>b</b>) Simplified spatial units of land use and road network; (<b>c</b>) Names of key Roads and spatial units.</p> Full article ">Figure 4
<p>Comparison of residents’ travels at different hours in different spatial units.</p> Full article ">Figure 5
<p>Real-time screenshots during model running and simulation. (<b>a</b>) Run time 30 s; (<b>b</b>) Run time 900 s; (<b>c</b>) Run time 1800 s; (<b>d</b>) Run time 3200 s.</p> Full article ">Figure 6
<p>Numbering and distribution of road intersections. (<b>a</b>) Numbering and distribution of existing road intersections; (<b>b</b>) Numbering and distribution of planned road intersections.</p> Full article ">Figure 7
<p>Changes in the number of agents at intersections.</p> Full article ">Figure 8
<p>Traffic forecasts for the four studied hours based on the big data from Gaode Map. (<b>a</b>) 7:00 traffic conditions; (<b>b</b>) 8:00 traffic conditions; (<b>c</b>) 9:00 traffic conditions; (<b>d</b>) 10:00 traffic conditions.</p> Full article ">Figure 9
<p>Comparison of traffic conditions before and after the optimization of the road network. (<b>a</b>) The traffic conditions of road intersections before optimization; (<b>b</b>) The traffic conditions of road intersections after optimization.</p> Full article ">
15 pages, 3130 KiB  
Article
Identifying Alternative Wetting and Drying (AWD) Adoption in the Vietnamese Mekong River Delta: A Change Detection Approach
by Robin J. Lovell
ISPRS Int. J. Geo-Inf. 2019, 8(7), 312; https://doi.org/10.3390/ijgi8070312 - 22 Jul 2019
Cited by 7 | Viewed by 5982
Abstract
Alternative wetting and drying (AWD) is an increasingly popular water-saving practice in rice production in the Vietnamese Mekong River Delta, especially considering the impact of projected climate change and reduced water availability. Unfortunately, it is very difficult to determine adoption without deploying thousands [...] Read more.
Alternative wetting and drying (AWD) is an increasingly popular water-saving practice in rice production in the Vietnamese Mekong River Delta, especially considering the impact of projected climate change and reduced water availability. Unfortunately, it is very difficult to determine adoption without deploying thousands of costly household surveys. This research used European Space Agency Sentinel-1a and 1b radar data, combined with in-situ moisture readings, to determine AWD adoption through change detection of a time series wetness index (WI). By using a beta coefficient of the radar data, the WI avoided the pitfalls of cloud cover, surface roughness, and vegetative interference that arise from the sigma coefficient data. The analysis illustrated an AWD adoption likelihood scale across the delta and it showed potential for the use of remotely sensed data to detect adoption. Trends across the Vietnamese delta showed higher adoption rates inland, with lower adoption of AWD in the coastal provinces. These results were supported by a simultaneous effort to collect household level adoption data as part of the same project. However, correlation between the WI values and in situ soil moisture meter readings were most accurate in alluvial soils, illustrating a particularly strong relationship between soil type and WI model robustness. The research suggests that future change detection efforts should focus on retrieving a multi-season dataset and employing a power density analysis on the time series data to fully understand the periodicity of dry down patterns. Full article
►▼ Show Figures

Figure 1

Figure 1
<p>Remote sensing workflow. The workflow was adapted and developed throughout the project, based on previous WI studies, as well as project iterations.</p> Full article ">Figure 2
<p>Study Location Map. Test plots and soil moisture meter locations are shown in red. Two test plots were chosen in each province, an alternative wetting and drying (AWD) and non-AWD rice paddy, representing the three major soil types in the delta. An Giang Province is alluvial soils, Ben Tre Province is saline, and Dong Thap Province is acid-sulfate. Note: the two plots are somewhat separated in An Giang, while they are very close together in Dong Thap and Ben Tre, so they display as a single point in the latter two provinces.</p> Full article ">Figure 3
<p>Change detection approach. (<b>a</b>) Initial time step wetness index (WI) scores for Test Plot 3, used in this figure as an example to illustrate the variability within each rice paddy, as well as its change over time. (<b>b</b>) The second time step in Test Plot 3, showing the wet and dry patterns changing over time. (<b>c</b>) Average change over time is shown, with bluer values illustrating a wetting trend and tan values illustrating a drying trend.</p> Full article ">Figure 4
<p>Plot 3 soil water content variability. The following are readings from Test Plot 3 in An Giang province, in which the farmer uses AWD management practices. It shows the VV-WI, and moisture percent readings for the plot. The two factors show a strong correlation (r = 0.65).</p> Full article ">Figure 5
<p>Plot 5 soil water content variability. The following readings from Test Plot 5 in An Giang province, in which the farmer uses flooded management practices, shows the VV WI and moisture percent readings for the plot. The two factors show a strong correlation (r = 0.59)</p> Full article ">Figure 6
<p>Inland average WI change. The inland pattern of higher dryness averages showing an overall greater occurrence of drying down, or a lower WI score. These results demonstrate that the inland provinces of An Giang and Dong Thap may be adopting AWD at a broader, more uniform rate than the coastal provinces.</p> Full article ">Figure 7
<p>Coastal average WI change. The coastal pattern of higher WI averages shows an overall average trend toward flooding or a higher WI score. These results demonstrate that coastal provinces, such as Ben Tre, adopt AWD practices less often than the inland provinces.</p> Full article ">
21 pages, 30093 KiB  
Article
COMET’s Education and Training for the Worldwide Meteorological Satellite User Community: Meeting Evolving Needs with Innovative Instruction
by Patrick Dills, Amy Stevermer, Tony Mancus, Bryan Guarente, Tim Alberta and Elizabeth Page
ISPRS Int. J. Geo-Inf. 2019, 8(7), 311; https://doi.org/10.3390/ijgi8070311 - 20 Jul 2019
Cited by 4 | Viewed by 4086
Abstract
Since 1989, the COMET<sup>&#xAE;</sup> Program&#x2019;s staff of instructional designers, scientists, graphic artists, and web developers has been creating targeted, effective, and scientifically sound instructional materials for the geosciences in multiple languages and formats. The majority of COMET training materials and services are available [...] Read more.
Since 1989, the COMET<sup>&#xAE;</sup> Program&#x2019;s staff of instructional designers, scientists, graphic artists, and web developers has been creating targeted, effective, and scientifically sound instructional materials for the geosciences in multiple languages and formats. The majority of COMET training materials and services are available via COMET&#x2019;s online training portal, MetEd. MetEd hosts over 500 self-paced English-language lessons, which are freely available to registered users. The lessons cover a broad array of topics, including satellite meteorology, numerical weather prediction, hydrometeorology, oceanography, aviation weather, climate science, and decision support. Nearly 300 lessons have been translated to other languages. NOAA NESDIS, EUMETSAT, the Meteorological Service of Canada, and the US National Weather Service all provide funding and subject matter expertise for satellite training efforts at COMET. The COMET team is focused on helping our sponsors refine their learning objectives and produce instructional material that is focused on learner engagement, knowledge retention, and measurable performance improvement. The COMET Program has continually transformed its instructional approach to better meet the shifting needs of learners. Our satellite remote sensing educational and training materials provide sound foundational knowledge for existing and new satellite products paired with increasing opportunities to apply that knowledge. Full article
(This article belongs to the Special Issue Education and Training in Applied Remote Sensing)
►▼ Show Figures

Graphical abstract

Graphical abstract
Full article ">Figure 1
<p>Map showing daily access to MetEd by country. MetEd serves over 190 countries and territories worldwide, with approximately 600,000 registered users as of April 2019.</p> Full article ">Figure 2
<p>Instrumentation and observation (which includes satellite topics) is the top training priority identified for WMO Regional Association (RA) IV (members include North America, Central America, Caribbean). Communication and customer interaction (key components of decision support services) is in the top five [<a href="#B4-ijgi-08-00311" class="html-bibr">4</a>].</p> Full article ">Figure 3
<p>The training development process design follows from Alessi and Trollip’s 2001 [<a href="#B1-ijgi-08-00311" class="html-bibr">1</a>] design and development model and incorporates the fundamental structure of the ADDIE (analysis, design, development, implementation, and evaluation) method combined with some elements of the successive approximation model (SAM) [<a href="#B10-ijgi-08-00311" class="html-bibr">10</a>].</p> Full article ">Figure 4
<p>Drawing tool activity where learners are asked to locate and outline locations for types of flooding discernible in satellite imagery. The top image displays learner input while the bottom image presents the correct locations (with white shadowing on the outlines) paired with the locations the learner has drawn for comparison. To view this interaction online, go to <a href="https://www.meted.ucar.edu/satmet/goes16_JPSS_hydro/navmenu.php?tab=1&amp;page=3-3-0&amp;type=flash" target="_blank">https://www.meted.ucar.edu/satmet/goes16_JPSS_hydro/navmenu.php?tab=1&amp;page=3-3-0&amp;type=flash</a>.</p> Full article ">Figure 5
<p>Learners can plot forecast model products and real-time satellite data using web map services enabled within a lesson to look at conditions for a weather situation in their region of interest (<a href="https://www.meted.ucar.edu/asmet/asmet11/EUMETSAT_NWP_Archive.htm" target="_blank">https://www.meted.ucar.edu/asmet/asmet11/EUMETSAT_NWP_Archive.htm</a>).</p> Full article ">Figure 6
<p>An interactive slider tool lets learners compare imagery products as they answer questions.</p> Full article ">Figure 7
<p>Screenshot of the MetEd education and training page, showing an example satellite training lesson that demonstrates land surface analysis using satellite products. On the right is a description of COMET’s lessons versus full distance learning courses.</p> Full article ">Figure 8
<p>Title page screenshots for four recent event-focused satellite training lessons on MetEd.</p> Full article ">Figure 9
<p>Cumulative number of users on MetEd from 2007 to present (April 2019). MetEd’s reach, both internationally and overall, has increased continuously, with approximately 600,000 learners currently registered on the site.</p> Full article ">Figure 10
<p>MetEd users can voluntarily report their affiliation when registering in the system. A total of 44% of MetEd learners are affiliated with universities, both within the US and internationally. Learners identifying as “weather enthusiasts” and “other” also represent a substantial number of users, and government sector forecasters and scientists make up other user groups on MetEd.</p> Full article ">Figure 11
<p>New satellite meteorology lessons published per year (blue) and cumulative number of satellite meteorology lessons (green) on MetEd from 1999 to March 2019. As of early April 2019, COMET offers approximately 120 satellite meteorology lessons on MetEd.</p> Full article ">Figure 12
<p>User session counts (through February 2019) for the eight SatFC-G lessons developed by COMET and published in 2016 prior to the GOES-R satellite launch. These lessons have received considerable use by meteorological services throughout the Americas, but particularly by NOAA’s US NWS.</p> Full article ">Figure 13
<p>The bar chart shows total user session counts for select microwave remote sensing lessons from 2007 to the present (April 2019). A session is defined as a learner accessing more than one page in the lesson, for at least one minute. Two of these lessons are now available as updated second editions.</p> Full article ">Figure 14
<p>The bar chart shows total user session counts for three frequently used satellite meteorology lessons on MetEd. The metrics are from publication date (shown following the title below each bar) through March 2019.</p> Full article ">Figure 15
<p>The bar graph displays the mean pre-assessment and post-lesson quiz performance (average percent out of 100) of learners who took both assessments for each of three case exercise lessons.</p> Full article ">
19 pages, 1000 KiB  
Article
Assessment and Benchmarking of Spatially Enabled RDF Stores for the Next Generation of Spatial Data Infrastructure
by Weiming Huang, Syed Amir Raza, Oleg Mirzov and Lars Harrie
ISPRS Int. J. Geo-Inf. 2019, 8(7), 310; https://doi.org/10.3390/ijgi8070310 - 19 Jul 2019
Cited by 20 | Viewed by 5041
Abstract
Geospatial information is indispensable for various real-world applications and is thus a prominent part of today’s data science landscape. Geospatial data is primarily maintained and disseminated through spatial data infrastructures (SDIs). However, current SDIs are facing challenges in terms of data integration and [...] Read more.
Geospatial information is indispensable for various real-world applications and is thus a prominent part of today’s data science landscape. Geospatial data is primarily maintained and disseminated through spatial data infrastructures (SDIs). However, current SDIs are facing challenges in terms of data integration and semantic heterogeneity because of their partially siloed data organization. In this context, linked data provides a promising means to unravel these challenges, and it is seen as one of the key factors moving SDIs toward the next generation. In this study, we investigate the technical environment of the support for geospatial linked data by assessing and benchmarking some popular and well-known spatially enabled RDF stores (RDF4J, GeoSPARQL-Jena, Virtuoso, Stardog, and GraphDB), with a focus on GeoSPARQL compliance and query performance. The tests were performed in two different scenarios. In the first scenario, geospatial data forms a part of a large-scale data infrastructure and is integrated with other types of data. In this scenario, we used ICOS Carbon Portal’s metadata—a real-world Earth Science linked data infrastructure. In the second scenario, we benchmarked the RDF stores in a dedicated SDI environment that contains purely geospatial data, and we used geospatial datasets with both crowd-sourced and authoritative data (the same test data used in a previous benchmark study, the Geographica benchmark). The assessment and benchmarking results demonstrate that the GeoSPARQL compliance of the RDF stores has encouragingly advanced in the last several years. The query performances are generally acceptable, and spatial indexing is imperative when handling a large number of geospatial objects. Nevertheless, query correctness remains a challenge for cross-database interoperability. In conclusion, the results indicate that the spatial capacity of the RDF stores has become increasingly mature, which could benefit the development of future SDIs. Full article
(This article belongs to the Special Issue SDI and the Revolutionary Technological Trends)
►▼ Show Figures

Figure 1

Figure 1
<p>Geographic locations of Integrated Carbon Observation System (ICOS) measurement stations.</p> Full article ">Figure 2
<p>Geospatial part of the ICOS metadata ontology. The concepts and relations without prefix annotation are from ICOS metadata ontology.</p> Full article ">
18 pages, 5705 KiB  
Article
VS30 Seismic Microzoning Based on a Geomorphology Map: Experimental Case Study of Chiang Mai, Chiang Rai, and Lamphun, Thailand
by Patcharavadee Thamarux, Masashi Matsuoka, Nakhorn Poovarodom and Junko Iwahashi
ISPRS Int. J. Geo-Inf. 2019, 8(7), 309; https://doi.org/10.3390/ijgi8070309 - 18 Jul 2019
Cited by 5 | Viewed by 4411
Abstract
Thailand is not known to be an earthquake-prone country; however, in 2014, an unexpected moderate earthquake caused severe damage to infrastructure and resulted in public panic. This event caught public attention and raised awareness of national seismic disaster management. However, the expertise and [...] Read more.
Thailand is not known to be an earthquake-prone country; however, in 2014, an unexpected moderate earthquake caused severe damage to infrastructure and resulted in public panic. This event caught public attention and raised awareness of national seismic disaster management. However, the expertise and primary data required for implementation of seismic disaster management are insufficient, including data on soil character which are used in amplification analyses for further ground motion prediction evaluations. Therefore, in this study, soil characterization was performed to understand the seismic responses of soil rigidity. The final output is presented in a seismic microzoning map. A geomorphology map was selected as the base map for the analysis. The geomorphology units were assigned with a time-averaged shear wave velocity of 30 m (VS30), which was collected by the spatial autocorrelation (SPAC) method of microtremor array measurements. The VS30 values were obtained from the phase velocity of the Rayleigh wave corresponding to a 40 m wavelength (C(40)). From the point feature, the VS30 values were transformed into polygonal features based on the geomorphological characteristics. Additionally, the automated geomorphology classification was explored in this study. Then, the seismic microzones were compared with the locations of major damage from the 2014 records for validation. The results from this study include geomorphological classification and seismic microzoning. The results suggest that the geomorphology units obtained from a pixel-based classification can be recommended for use in seismic microzoning. For seismic microzoning, the results show mainly stiff soil and soft rocks in the study area, and these geomorphological units have relatively high amplifications. The results of this study provide a valuable base map for further disaster management analyses. Full article
(This article belongs to the Special Issue Geomatics and Geo-Information in Earthquake Studies)
►▼ Show Figures

Figure 1

Figure 1
<p>Study area: the districts of Chiang Mai, Chiang Rai, and Lamphun with the V<sub>S30</sub> observation stations.</p> Full article ">Figure 2
<p>Digital elevation model (DEM) scenes for geomorphology classification: (<b>i</b>) AOI1 covers Asia–Oceania (resolution: 1000 m) and (<b>ii</b>) AOI2 covers Chiang Mai, Chiang Rai, and Lamphun. (resolution: 250 m).</p> Full article ">Figure 3
<p>Major geomorphology groups.</p> Full article ">Figure 4
<p>Descriptions of classified geomorphology units from pixel-based classification (PBC) and object-based classification (OBC) [<a href="#B9-ijgi-08-00309" class="html-bibr">9</a>,<a href="#B50-ijgi-08-00309" class="html-bibr">50</a>].</p> Full article ">Figure 5
<p>Sample of array observation station.</p> Full article ">Figure 6
<p>Geomorphology map for AOI2 obtained using a 250-m resolution DEM (<a href="#ijgi-08-00309-f002" class="html-fig">Figure 2</a>(ii)).</p> Full article ">Figure 7
<p>V<sub>S30</sub> based on PBC geomorphological units; the red lines indicate the National Earthquake Hazards Reduction Program (NEHRP) classes.</p> Full article ">Figure 8
<p>V<sub>S30</sub> based on OBC geomorphological units; the red lines indicate the NEHRP classes.</p> Full article ">Figure 9
<p>V<sub>S30</sub> map based on the reclassified geomorphological map with NEHRP soil characteristic standards.</p> Full article ">Figure 10
<p>(<b>i</b>) Geology map of Thailand [<a href="#B61-ijgi-08-00309" class="html-bibr">61</a>]: the geological units are fluvial deposit (flood plain [Qff]), terrace and colluvium [Qt], alluvium (Qa), rhyolite [PTrv], semiconsolidate [Tmm], and granite [Trgr]; (<b>ii</b>) OBC geomorphology map from Iwahashi et al. [<a href="#B50-ijgi-08-00309" class="html-bibr">50</a>]; and (<b>iii</b>) classified PBC geomorphology map. The geomorphological units for(<b>ii</b>,<b>iii</b>) are based on the descriptions in <a href="#ijgi-08-00309-f004" class="html-fig">Figure 4</a>.</p> Full article ">Figure 11
<p>Zoomed-in hazard microzoning map with damage locations in 2014.</p> Full article ">
20 pages, 3645 KiB  
Article
Anomalous Urban Mobility Pattern Detection Based on GPS Trajectories and POI Data
by Zhenzhou Xu, Ge Cui, Ming Zhong and Xin Wang
ISPRS Int. J. Geo-Inf. 2019, 8(7), 308; https://doi.org/10.3390/ijgi8070308 - 17 Jul 2019
Cited by 24 | Viewed by 5139
Abstract
Anomalous urban mobility pattern refers to abnormal human mobility flow in a city. Anomalous urban mobility pattern detection is important in the study of urban mobility. In this paper, a framework is proposed to identify anomalous urban mobility patterns based on taxi GPS [...] Read more.
Anomalous urban mobility pattern refers to abnormal human mobility flow in a city. Anomalous urban mobility pattern detection is important in the study of urban mobility. In this paper, a framework is proposed to identify anomalous urban mobility patterns based on taxi GPS trajectories and Point of Interest (POI) data. In the framework, functional regions are first generated based on the distribution of POIs by the DBSCAN clustering algorithm. A Weighted Term Frequency-Inverse Document Frequency (WTF-IDF) method is proposed to identify function values in each region. Then, the Origin-Destination (OD) of trips between functional regions is extracted from GPS trajectories to detect anomalous urban mobility patterns. Mobility vectors are established for each time interval based on the OD of trips and are classified into clusters by the mean shift algorithm. Abnormal urban mobility patterns are identified by processing the mobility vectors. A case study in the city of Wuhan, China, is conducted; the experimental results show that the proposed method can effectively identify daily and hourly anomalous urban mobility patterns. Full article
►▼ Show Figures

Figure 1

Figure 1
<p>Framework of anomalous mobility pattern detection.</p> Full article ">Figure 2
<p>The spatial distribution of POIs. (<b>a</b>) Along the street; (<b>b</b>) concentrated around a center.</p> Full article ">Figure 3
<p>Top 10 intensive functional regions.</p> Full article ">Figure 4
<p>Examples of the function distribution in functional regions. (<b>a</b>) Region 155; (<b>b</b>) Region 156.</p> Full article ">Figure 5
<p>Temporal distribution of Origin-Destination (OD) trips.</p> Full article ">Figure 6
<p>The overall view of urban mobility in Wuhan.</p> Full article ">Figure 7
<p>The similarity matrix of mobility vectors at the scale of day.</p> Full article ">Figure 8
<p>The top 10 functional regions with the largest values in (<b>a</b>) <math display="inline"><semantics> <mrow> <msub> <mi>V</mi> <mn>2</mn> </msub> </mrow> </semantics></math> (<b>b</b>) <math display="inline"><semantics> <mrow> <msub> <mi>V</mi> <mn>3</mn> </msub> </mrow> </semantics></math> and (<b>c</b>) <math display="inline"><semantics> <mrow> <msub> <mi>V</mi> <mn>4</mn> </msub> </mrow> </semantics></math>.</p> Full article ">
24 pages, 7617 KiB  
Article
Monitoring Ground Instabilities Using SAR Satellite Data: A Practical Approach
by Matteo Del Soldato, Lorenzo Solari, Federico Raspini, Silvia Bianchini, Andrea Ciampalini, Roberto Montalti, Alessandro Ferretti, Vania Pellegrineschi and Nicola Casagli
ISPRS Int. J. Geo-Inf. 2019, 8(7), 307; https://doi.org/10.3390/ijgi8070307 - 17 Jul 2019
Cited by 47 | Viewed by 6101
Abstract
Satellite interferometric data are widely exploited for ground motion monitoring thanks to their wide area coverage, cost efficiency and non-invasiveness. The launch of the Sentinel-1 constellation opened new horizons for interferometric applications, allowing the scientists to rethink the way in which these data [...] Read more.
Satellite interferometric data are widely exploited for ground motion monitoring thanks to their wide area coverage, cost efficiency and non-invasiveness. The launch of the Sentinel-1 constellation opened new horizons for interferometric applications, allowing the scientists to rethink the way in which these data are delivered, passing from a static view of the territory to a continuous streaming of ground motion measurements from space. Tuscany Region is the first worldwide example of a regional scale monitoring system based on satellite interferometric data. The processing chain here exploited combines a multi-interferometric approach with a time-series data mining algorithm aimed at recognizing benchmarks with significant trend variations. The system is capable of detecting the temporal changes of a wide variety of phenomena such as slow-moving landslides and subsidence, producing a high amount of data to be interpreted in a short time. Bulletins and reports are derived to the hydrogeological risk management actors at regional scale. The final output of the project is a list of potentially hazardous and accelerating phenomena that are verified on site by field campaign by completing a sheet survey in order to qualitatively estimate the risk and to suggest short-term actions to be taken by local entities. Two case studies, one related to landslides and one to subsidence, are proposed to highlight the potential of the monitoring system to early detect anomalous ground changes. Both examples represent a successful implementation of satellite interferometric data as monitoring and risk management tools, raising the awareness of local and regional authorities to geohazards. Full article
(This article belongs to the Special Issue Geospatial Approaches to Landslide Mapping and Monitoring)
►▼ Show Figures

Graphical abstract

Graphical abstract
Full article ">Figure 1
<p>Geographical (<b>a</b>) and geomorphological (<b>b</b>) framework of Tuscany Region. The background of inset (<b>a</b>) is an aerial orthophoto referred to year 2012. The Digital Terrain Model (<b>b</b>) has a resolution of 10 m.</p> Full article ">Figure 2
<p>Work flow approach for managing systematic updated Sentinel-1 InSAR data at regional scale.</p> Full article ">Figure 3
<p>Schematic explanation of the automatic APs detection algorithm. Δt is the temporal search window, Δv is the velocity variation. The blue points represent a possible time series without variations, whereas the red points simulate an abrupt acceleration.</p> Full article ">Figure 4
<p>Structure of the monitoring bulletin. Front page with municipality classification (<b>a</b>), information about the AP (<b>b</b>), deformation map of the area with relevant and persistent APs (<b>c</b>) and time-series (<b>d</b>).</p> Full article ">Figure 5
<p>Flowchart of the field survey to collect all the data for combining the Intensity (I) and Exposure (E) for the preliminary risk (R) evaluation.</p> Full article ">Figure 6
<p>Ascending (<b>a</b>) and descending (<b>b</b>) deformation maps obtained by the SqueeSAR processing of Sentinel-1 images. Movements towards the satellite are highlighted in cold colours, whereas displacements away from the satellite are displayed in hot colours. The stability range is equal to ±2 mm/yr (green points).</p> Full article ">Figure 7
<p>Classification of anomalous points according to the causes (<b>a</b>). On the right, an explanation of the presence of APs divided by causes for the entire region (<b>b</b>) and extrapolated for each province (<b>c</b>) (updated after [<a href="#B53-ijgi-08-00307" class="html-bibr">53</a>]).</p> Full article ">Figure 8
<p>Multi-temporal bulletins delivered during the project (<b>a</b>) and all the “red” municipalities reported during the two years of continuous monitoring (<b>b</b>).</p> Full article ">Figure 9
<p>Localization and deformation map of the area of Carpineta landslide (<b>a</b>). Inset (<b>b</b>) shows 2 time series of MP without trend changes (blue) and with trend change higher than 10 mm/yr (red).</p> Full article ">Figure 10
<p>Cracks recognized during the field survey in the Carpineta hamlet. Landslide induced damage recognizable for the external walls (Photo 1, 2 and 3).</p> Full article ">Figure 11
<p>Localization and deformation map of the area of Montemurlo (<b>a</b>) with 2 time series of ascending (blue) and descending (red) APs showing two abrupt trend changes (black lines) (<b>b</b>).</p> Full article ">Figure 12
<p>Cracks recognized during the field survey in Montemurlo. Subsidence-induced damage recognizable in the internal (Photo 1 and 2) structures of a textile shed. The same structure records a small tilting of few degrees (see the concrete column with respect to the green wall of the shed—Photo 3).</p> Full article ">
14 pages, 3657 KiB  
Article
Modeling the Vagueness of Areal Geographic Objects: A Categorization System
by Yu Liu, Yihong Yuan and Song Gao
ISPRS Int. J. Geo-Inf. 2019, 8(7), 306; https://doi.org/10.3390/ijgi8070306 - 17 Jul 2019
Cited by 13 | Viewed by 4086
Abstract
Modeling vague objects with indeterminate boundaries has drawn much attention in geographic information science. Because fields and objects are two perspectives in modeling geographic phenomena, this paper investigates the characteristics of vague regions from the perspective of the field/object dichotomy. Based on the [...] Read more.
Modeling vague objects with indeterminate boundaries has drawn much attention in geographic information science. Because fields and objects are two perspectives in modeling geographic phenomena, this paper investigates the characteristics of vague regions from the perspective of the field/object dichotomy. Based on the assumption that a vague object can be viewed as the conceptualization of a field, we defined five categories of vague objects: direct field-cutting objects, focal operation-based field-cutting objects, element-clustering objects, object-referenced objects, and dynamic boundary objects. We then established a categorization system to formalize the semantic differences between vague objects using the fuzzy set theory. The proposed framework provides valuable input for the conceptualization, interpretation, and modeling of vague geographical objects. Full article
►▼ Show Figures

Figure 1

Figure 1
<p>A three-phased conceptualization for woodland objects, using the field model.</p> Full article ">Figure 2
<p>The dilemma in conceptualizing a woodland. (<b>a</b>) A single tree is clearly not a piece of woodland. (<b>b</b>) A woodland consists of a number of trees with a certain density. (<b>c</b>) Whether a tree belongs to a woodland depends on the distance between this particular tree and other trees.</p> Full article ">Figure 3
<p>(<b>a</b>) The Tibetan Plateau and (<b>b</b>) its membership function.</p> Full article ">Figure 4
<p>The south slope area of (<b>a</b>) a volcano and (<b>b</b>) its membership function.</p> Full article ">Figure 5
<p>(<b>a</b>) Hypothetical ECOs based on Gestalt psychology; (<b>b</b>) Woodland: an example of ECOs.</p> Full article ">Figure 6
<p>An example of ORO: northern and southern California. (<b>a</b>) Vague regions purely based on spatial relations; (<b>b</b>) Vague cognitive regions using social media data.</p> Full article ">Figure 7
<p>Difference between ECOs (<b>left</b>) and OROs (<b>right</b>), represented by UML class diagrams.</p> Full article ">Figure 8
<p>A lake as an example of DBO.</p> Full article ">Figure 9
<p>The categorization system of the five categories of fuzzy regions and their relations.</p> Full article ">
23 pages, 3425 KiB  
Article
Assessing Spatial Information Themes in the Spatial Information Infrastructure for Participatory Urban Planning Monitoring: Indonesian Cities
by Agung Indrajit, Bastiaan Van Loenen and Peter Van Oosterom
ISPRS Int. J. Geo-Inf. 2019, 8(7), 305; https://doi.org/10.3390/ijgi8070305 - 17 Jul 2019
Cited by 9 | Viewed by 5267
Abstract
Most urban planning monitoring activities were designed to monitor implementation of aggregated sectors from different initiatives into practical and measurable indicators. Today, cities utilize spatial information in monitoring and evaluating urban planning implementation for not only national or local goals but also for [...] Read more.
Most urban planning monitoring activities were designed to monitor implementation of aggregated sectors from different initiatives into practical and measurable indicators. Today, cities utilize spatial information in monitoring and evaluating urban planning implementation for not only national or local goals but also for the 2030 Agenda of Sustainable Development Goals (SDGs). Modern cities adopt Participatory Geographic Information System (PGIS) initiative for their urban planning monitoring. Cities provide spatial information and online tools for citizens to participate. However, the selection of spatial information services for participants is made from producers’ perception and often disregards requirements from the regulation, functionalities, and broader user’s perception. By providing appropriate spatial information, the quality of participatory urban monitoring can be improved. This study presents a method for selecting appropriate spatial information for urban planning monitoring. It considers regulation, urban planning, and spatial science theories, as well as citizens’ requirements, to support participatory urban planning monitoring as a way to ensure the success of providing near real-time urban information to planners and decision-makers. Full article
(This article belongs to the Special Issue Algorithms and Techniques in Urban Monitoring)
►▼ Show Figures

Graphical abstract

Graphical abstract
Full article ">Figure 1
<p>The McLoughlin implementation process. The planning process model interacts with the real world through comparative analyses and control processes [<a href="#B16-ijgi-08-00305" class="html-bibr">16</a>].</p> Full article ">Figure 2
<p>Methodology for determination of requirements for spatial information services for Participatory Urban Planning Monitoring adapted from Malinowski and Zimányi [<a href="#B14-ijgi-08-00305" class="html-bibr">14</a>].</p> Full article ">Figure 3
<p>Detailed Urban Planning in Indonesian Spatial Planning System according to the Minister of Public Work Decree 20 Year 2011 [<a href="#B67-ijgi-08-00305" class="html-bibr">67</a>].</p> Full article ">Figure 4
<p>Relationship between Urban Planning, Detailed Urban Planning, and Building and Environment Planning according to the Minister of Public Work Decree 20 Year 2011 [<a href="#B67-ijgi-08-00305" class="html-bibr">67</a>].</p> Full article ">Figure 5
<p>The workflow of Urban Planning Monitoring Process in Indonesia according to Indonesian Government Regulation No. 15 Year 2010 [<a href="#B64-ijgi-08-00305" class="html-bibr">64</a>].</p> Full article ">Figure 6
<p>The workflow of conformance checking for urban planning monitoring derived from the Indonesian Government Regulation No. 15 Year 2010 [<a href="#B64-ijgi-08-00305" class="html-bibr">64</a>].</p> Full article ">Figure 7
<p>Respondents based on affiliation (sectors).</p> Full article ">Figure 8
<p>Respondents based on actor role derived from their affiliation.</p> Full article ">Figure 9
<p>Type of spatial information needed by urban planners, citizens, and non-government institutions to contribute to participatory urban planning monitoring.</p> Full article ">Figure 10
<p>The quality of spatial information expected by urban planners, citizens, and non-government institutions in participatory urban planning monitoring.</p> Full article ">Figure 11
<p>Preferences of spatial information to be used in participatory urban planning monitoring from the potential contributor (<b>red</b>) and urban planners (<b>blue</b>) (in percentages).</p> Full article ">
23 pages, 4400 KiB  
Article
A New Algorithms of Stroke Generation Considering Geometric and Structural Properties of Road Network
by Yi Liu and Wenjing Li
ISPRS Int. J. Geo-Inf. 2019, 8(7), 304; https://doi.org/10.3390/ijgi8070304 - 16 Jul 2019
Cited by 10 | Viewed by 3808
Abstract
Strokes are considered an elementary unit of road networks and have been widely used in their analysis and application. However, most conventional stroke generation methods are based solely on a fixed angle threshold, which ignores road networks’ geometric and structural properties. To remedy [...] Read more.
Strokes are considered an elementary unit of road networks and have been widely used in their analysis and application. However, most conventional stroke generation methods are based solely on a fixed angle threshold, which ignores road networks’ geometric and structural properties. To remedy this, this paper proposes an algorithm for generating strokes that takes into account these additional geometric and structural road network properties and that reduces the impact of stroke generation on road network quality. To this end, we introduce a model of feature-based information entropy and then utilize this model to calculate road networks’ information volume and both the elemental and neighborhood level. To make our experimental results more objective, we use the Douglas-Peucker algorithm to simplify the information change curve and to obtain the optimal angle threshold range for generating strokes for different road network structures. Finally, we apply this model to three different road networks, and the optimal threshold ranges are 54°–63° (Chicago), 61°–63° (Moscow), 45°–48° (Monaco). And taking Monaco as an example, this paper conducts stroke selection experiments. The results demonstrate that our proposed algorithm has better connectivity and wider coverage than those based on a common angle threshold (60°). Full article
(This article belongs to the Special Issue Map Generalization)
►▼ Show Figures

Figure 1

Figure 1
<p>Flow chart of our model for determining the optimal angle threshold range for generating strokes in a specific road network.</p> Full article ">Figure 2
<p>Calculation of road deflection angle. θ is the deflection angle of road segments R1 and R2.</p> Full article ">Figure 3
<p>Illustration of different connection strategies.</p> Full article ">Figure 4
<p>Flow chart for our improved stroke-determination algorithm.</p> Full article ">Figure 5
<p>Stroke generation using different angle thresholds.</p> Full article ">Figure 6
<p>Wuhan road network at different levels.</p> Full article ">Figure 7
<p>Simplified results of information volume changing curve. The numbers represent the slope at a given point.</p> Full article ">Figure 7 Cont.
<p>Simplified results of information volume changing curve. The numbers represent the slope at a given point.</p> Full article ">Figure 8
<p>Case study areas, both original road network and dual graph.</p> Full article ">Figure 9
<p>The information volume curve of each road network.</p> Full article ">Figure 10
<p>Monaco’s optimal angle threshold according to Douglas-Peucker algorithm.</p> Full article ">Figure 11
<p>Chicago’s optimal angle threshold according to Douglas-Peucker algorithm.</p> Full article ">Figure 12
<p>Moscow’s optimal angle threshold according to Douglas-Peucker algorithm.</p> Full article ">Figure 12 Cont.
<p>Moscow’s optimal angle threshold according to Douglas-Peucker algorithm.</p> Full article ">Figure 13
<p>Results of stroke selection.</p> Full article ">Figure 13 Cont.
<p>Results of stroke selection.</p> Full article ">
13 pages, 4743 KiB  
Article
Skeleton Line Extraction Method in Areas with Dense Junctions Considering Stroke Features
by Chengming Li, Yong Yin, Pengda Wu and Wei Wu
ISPRS Int. J. Geo-Inf. 2019, 8(7), 303; https://doi.org/10.3390/ijgi8070303 - 16 Jul 2019
Cited by 8 | Viewed by 3201
Abstract
Extraction of the skeleton line of complex polygons is difficult, and a hot topic in map generalization study. Due to the irregularity and complexity of junctions, it is difficult for traditional methods to maintain main structure and extension characteristics when dealing with dense [...] Read more.
Extraction of the skeleton line of complex polygons is difficult, and a hot topic in map generalization study. Due to the irregularity and complexity of junctions, it is difficult for traditional methods to maintain main structure and extension characteristics when dealing with dense junction areas, so a skeleton line extraction method considering stroke features has been proposed in this paper. Firstly, we put forward a long-edge adaptive node densification algorithm, which is used to construct boundary-constrained Delaunay triangulation to uniformly divide the polygon and extract the initial skeleton line. Secondly, we defined the triangles with three adjacent triangles (Type III) as the basic unit of junctions, then obtained the segmented areas with dense junctions on the basis of local width characteristics and correlation relationships of each Type III triangle. Finally, we concatenated the segments into strokes and corrected the initial skeleton lines based on the extension direction features of each stroke. The actual water network data of Jiangsu Province in China were used to verify the method. Experimental results show that the proposed method can better identify the areas with dense junctions and that the extracted skeleton line is naturally smooth and well-connected, which accurately reflects the main structure and extension characteristics of these areas. Full article
►▼ Show Figures

Figure 1

Figure 1
<p>Triangle classification: (<b>a</b>) Type I triangle; (<b>b</b>) Type II triangle; (<b>c</b>) Type III triangle.</p> Full article ">Figure 2
<p>Triangle classification: (<b>a</b>) Original data with dense junctions; (<b>b</b>) the skeleton lines (green solid line) extracted by existing method.</p> Full article ">Figure 3
<p>The process diagram for our method</p> Full article ">Figure 4
<p>Long-edge adaptive node densification: (<b>a</b>) Stretched Type III triangles (blue); (<b>b</b>) Adjacent Type II triangle (yellow); (<b>c</b>) The longest edge in Type II (purple); (<b>d</b>) Densifying nodes on the long-edge.</p> Full article ">Figure 5
<p>Junction areas association judgment: (<b>a</b>) Junction nodes; (<b>b</b>) Effective length <span class="html-italic">Lv</span> of connecting arc.</p> Full article ">Figure 6
<p>The stroke in areas with dense junctions (the thick blue line).</p> Full article ">Figure 7
<p>Skeleton line adjustment in dense junction area: (<b>a</b>) Original data; (<b>b</b>) Adjustment result of the area with branch in an unstable direction; (<b>c</b>) Adjustment result of the area with branch in a stable direction.</p> Full article ">Figure 8
<p>Comparison of extraction results of skeleton lines in simple junction areas: (<b>a</b>) Original skeleton line; (<b>b</b>) Skeleton line of Li [<a href="#B17-ijgi-08-00303" class="html-bibr">17</a>]; (<b>c</b>) Skeleton line of this paper.</p> Full article ">Figure 9
<p>Comparison of extraction results of skeleton lines in complex junction areas: (<b>a</b>) Original skeleton line; (<b>b</b>) Skeleton line of Li [<a href="#B17-ijgi-08-00303" class="html-bibr">17</a>]; (<b>c</b>) Skeleton line of this paper.</p> Full article ">Figure 10
<p>Comparison of extraction results of skeleton lines for the sample data in <a href="#ijgi-08-00303-f002" class="html-fig">Figure 2</a>a: (<b>a</b>) Skeleton line of Li [<a href="#B17-ijgi-08-00303" class="html-bibr">17</a>]; (<b>b</b>) Skeleton line of this paper.</p> Full article ">
13 pages, 1232 KiB  
Article
Areal Interpolation Using Parcel and Census Data in Highly Developed Urban Environments
by XiaoHang Liu and Alexis Martinez
ISPRS Int. J. Geo-Inf. 2019, 8(7), 302; https://doi.org/10.3390/ijgi8070302 - 16 Jul 2019
Cited by 8 | Viewed by 4426
Abstract
Areal interpolation is routinely used when spatial data are unavailable at desired geographical units. While many methods are available, few of them were developed specifically for and tested in highly developed urban cores. Even fewer studied subpopulation or population characteristics. This paper explores [...] Read more.
Areal interpolation is routinely used when spatial data are unavailable at desired geographical units. While many methods are available, few of them were developed specifically for and tested in highly developed urban cores. Even fewer studied subpopulation or population characteristics. This paper explores both issues using parcel map and decennial census data as ancillary information. Using census blocks as intermediate zones, the method first disaggregates source-zone data to intermediate zones, then disaggregates data to parcel level in intermediate zones intersecting target zones, and finally aggregates intermediate-zone and parcel-level estimates to obtain target-zone estimates. Compared to areal weighting and residential proportion, the proposed method is significantly more accurate. All three methods perform the best on population count, and worst on spatially clustered subpopulations such as black/African American population. Quotient variables are more difficult to interpolate than count variables. The research demonstrates the utility of parcel and decennial census data for areal interpolation in highly developed urban cores, and calls for future research on subpopulation and population characteristics. Full article
►▼ Show Figures

Figure 1

Figure 1
<p>The study area.</p> Full article ">Figure 2
<p>An example where areal weighting will result in poor estimates.</p> Full article ">Figure 3
<p>Areal interpolation using three methods: (<b>a</b>) areal interpolation, (<b>b</b>) residential proportion, and (<b>c</b>) parcel and decennial census data as ancillary information.</p> Full article ">Figure 4
<p>A case where the use of decennial census data results in improved accuracy. Parcel data are used to allocate 285 black residents to the blocks, with and without decennial census data. A label such as ‘132/47/191’ means the block has 132 black residents in 2010; it is estimated to have 47 black residents in 2013 when decennial data are used, and 191 when not.</p> Full article ">
17 pages, 5641 KiB  
Article
HBIM Modeling from the Surface Mesh and Its Extended Capability of Knowledge Representation
by Xiucheng Yang, Yi-Chou Lu, Arnadi Murtiyoso, Mathieu Koehl and Pierre Grussenmeyer
ISPRS Int. J. Geo-Inf. 2019, 8(7), 301; https://doi.org/10.3390/ijgi8070301 - 15 Jul 2019
Cited by 74 | Viewed by 7562
Abstract
Built heritage has been documented by reality-based modeling for geometric description and by ontology for knowledge management. The current challenge still involves the extraction of geometric primitives and the establishment of their connection to heterogeneous knowledge. As a recently developed 3D information modeling [...] Read more.
Built heritage has been documented by reality-based modeling for geometric description and by ontology for knowledge management. The current challenge still involves the extraction of geometric primitives and the establishment of their connection to heterogeneous knowledge. As a recently developed 3D information modeling environment, building information modeling (BIM) entails both graphical and non-graphical aspects of the entire building, which has been increasingly applied to heritage documentation and generates a new issue of heritage/historic BIM (HBIM). However, HBIM needs to additionally deal with the heterogeneity of geometric shape and semantic knowledge of the heritage object. This paper developed a new mesh-to-HBIM modeling workflow and an integrated BIM management system to connect HBIM elements and historical knowledge. Using the St-Pierre-le-Jeune Church, Strasbourg, France as a case study, this project employs Autodesk Revit as a BIM environment and Dynamo, a built-in visual programming tool of Revit, to extend the new HBIM functions. The mesh-to-HBIM process segments the surface mesh, thickens the triangle mesh to 3D volume, and transfers the primitives to BIM elements. The obtained HBIM is then converted to the ontology model to enrich the heterogeneous knowledge. Finally, HBIM geometric elements and ontology semantic knowledge is joined in a unified BIM environment. By extending the capability of the BIM platform, the HBIM modeling process can be conducted in a time-saving way, and the obtained HBIM is a semantic model with object-oriented knowledge. Full article
(This article belongs to the Special Issue BIM for Cultural Heritage (HBIM))
►▼ Show Figures

Figure 1

Figure 1
<p>Workflows of the HBIM modeling and the integration with ontological knowledge.</p> Full article ">Figure 2
<p>The main façade (left), cleaned point clouds (middle), and mesh geometry (right) of the St-Pierre-le-Jeune Church in Strasbourg, France, built from typical Alsatian red sandstone.</p> Full article ">Figure 3
<p>(<b>a</b>) The point clouds imported in BIM platform, (<b>b</b>) the HBIM structures created by reverse design, and (<b>c</b>) some typical “classes” composing the church on the reference of the point cloud</p> Full article ">Figure 4
<p>The explosion processing to segment the holistic mesh (<b>a</b>) into small blocks by a multi-layered explosion (<b>b</b>,<b>c</b>). The primitive representing the church’s primary component is then generated based on the merging of several small blocks (<b>d</b>). The surface mesh (<b>e</b>) is converted to NURBS curve based solid geometry by adding thickness (<b>f</b>) and translated in a BIM environment (<b>g</b>).</p> Full article ">Figure 5
<p>The segmentation, combination and extrusion process to transfer the holistic surface mesh into individual solid elements</p> Full article ">Figure 6
<p>The correspondence between HBIM and ontology.</p> Full article ">Figure 7
<p>Ontological knowledge modeling for the HBIM entities within the standards of CIDOC CRM.</p> Full article ">Figure 8
<p>Object property of the ontological knowledge model.</p> Full article ">Figure 9
<p>The connection of HBIM elements and ontological knowledge based on <span class="html-italic">Revit Dynamo,</span> where the corresponding knowledge of the element selected in <span class="html-italic">Revit</span> can be displayed in the “<span class="html-italic">watch</span>” window.</p> Full article ">Figure 10
<p>The final HBIM models in 2D and 3D views using manual scan-to-HBIM (left and middle) and semi-automated mesh-to-HBIM (right).</p> Full article ">Figure 11
<p>The browsing result of some heritage components in <span class="html-italic">Dynamo</span> with geometry and knowledge. The “<span class="html-italic">watch</span>” window (<b>b</b>, <b>d</b>, <b>e</b>, <b>f</b>) in the <span class="html-italic">Dynamo</span> interface can display the knowledge of selected element of HBIM model (<b>a</b>) in the <span class="html-italic">Revit</span> platform. The knowledge in “<span class="html-italic">watch</span>” window derives from previously defined knowledge model (<b>c</b>).</p> Full article ">
25 pages, 13682 KiB  
Article
A Convolutional Neural Network Architecture for Auto-Detection of Landslide Photographs to Assess Citizen Science and Volunteered Geographic Information Data Quality
by Recep Can, Sultan Kocaman and Candan Gokceoglu
ISPRS Int. J. Geo-Inf. 2019, 8(7), 300; https://doi.org/10.3390/ijgi8070300 - 15 Jul 2019
Cited by 60 | Viewed by 7331
Abstract
Several scientific processes benefit from Citizen Science (CitSci) and VGI (Volunteered Geographical Information) with the help of mobile and geospatial technologies. Studies on landslides can also take advantage of these approaches to a great extent. However, the quality of the collected data by [...] Read more.
Several scientific processes benefit from Citizen Science (CitSci) and VGI (Volunteered Geographical Information) with the help of mobile and geospatial technologies. Studies on landslides can also take advantage of these approaches to a great extent. However, the quality of the collected data by both approaches is often questionable, and automated procedures to check the quality are needed for this purpose. In the present study, a convolutional neural network (CNN) architecture is proposed to validate landslide photos collected by citizens or nonexperts and integrated into a mobile- and web-based GIS environment designed specifically for a landslide CitSci project. The VGG16 has been used as the base model since it allows finetuning, and high performance could be achieved by selecting the best hyper-parameters. Although the training dataset was small, the proposed CNN architecture was found to be effective as it could identify the landslide photos with 94% precision. The accuracy of the results is sufficient for purpose and could even be improved further using a larger amount of training data, which is expected to be obtained with the help of volunteers. Full article
(This article belongs to the Special Issue Advance Geospatial Artificial Intelligence for Landslide Modeling, Prediction and Management)
►▼ Show Figures

Figure 1

Figure 1
<p>General workflow of the study.</p> Full article ">Figure 2
<p>Examples of the photos used for training of each class type.</p> Full article ">Figure 3
<p>Two examples of the preprocessing and augmentation of the images. (<b>a</b>,<b>b</b>) are the original images, and the square photos are the cropped and augmented ones.</p> Full article ">Figure 4
<p>The original VGG16 [<a href="#B55-ijgi-08-00300" class="html-bibr">55</a>] architecture (<b>left</b>) and the proposed architecture of this study (<b>right</b>).</p> Full article ">Figure 5
<p>The first step of training in the proposed algorithm.</p> Full article ">Figure 6
<p>The second (final) step of the training in the proposed algorithm.</p> Full article ">Figure 7
<p>Training loss and accuracy results for the ten different runs of the model.</p> Full article ">Figure 7 Cont.
<p>Training loss and accuracy results for the ten different runs of the model.</p> Full article ">Figure 8
<p>Normalized confusion matrix obtained from the forth test run (Train–Test Split 4 model).</p> Full article ">
14 pages, 3748 KiB  
Article
Collision Detection for UAVs Based on GeoSOT-3D Grids
by Weixin Zhai, Xiaochong Tong, Shuangxi Miao, Chengqi Cheng and Fuhu Ren
ISPRS Int. J. Geo-Inf. 2019, 8(7), 299; https://doi.org/10.3390/ijgi8070299 - 15 Jul 2019
Cited by 15 | Viewed by 4440
Abstract
The increasing number of unmanned aerial vehicles (UAVs) has led to challenges related to solving the collision problem to ensure air traffic safety. The traditional approaches employed for collision detection suffer from two main drawbacks: first, the computational burden of a pairwise calculation [...] Read more.
The increasing number of unmanned aerial vehicles (UAVs) has led to challenges related to solving the collision problem to ensure air traffic safety. The traditional approaches employed for collision detection suffer from two main drawbacks: first, the computational burden of a pairwise calculation increases exponentially with an increasing number of spatial entities; second, existing grid-based approaches are unsuitable for complicated scenarios with a large number of objects moving at high speeds. In the proposed model, we first identified UAVs and other spatial objects with GeoSOT-3D grids. Second, the nonrelational spatial database was initialized with a multitable strategy, and spatiotemporal data were inserted with the GeoSOT-3D grid codes as the primary key. Third, the collision detection procedure was transformed from a pairwise calculation to a multilevel query. Four simulation experiments were conducted to verify the feasibility and efficiency of the proposed collision detection model for UAVs in different environments. The results also indicated that 64 m GeoSOT-3D grids are the most suitable basic grid size, and the reduction in the time consumption compared with traditional methods reached approximately 50–80% in different scenarios. Full article
(This article belongs to the Special Issue Global Grid Systems)
►▼ Show Figures

Figure 1

Figure 1
<p>Detailed illustration of GeoSOT-3D grids on Level 2. The codes for the three-dimensional grids are marked on the surface of the cube and on four separate layers. The binary codes are also listed for the front side, right side, and each layer.</p> Full article ">Figure 2
<p>Traditional calculation method and spatial database-based query method for unmanned aerial vehicle (UAV) collision detection. Traditional collision detection methods depend on pairwise calculations based on three-dimensional coordinates (<b>a</b>). The calculation complexity increases rapidly as the number of UAVs increases. The grid-based method simplifies the collision detection procedure as a comparison process for regular-sized spatial grids (<b>b</b>). The grid-based method can be further advanced with a UAV management table (<b>c</b>). In the table, the GeoSOT-3D code is set as the primary key, and the collision can be detected by a straightforward query.</p> Full article ">Figure 3
<p>Multitable strategy. There are a number of data tables in the nonrelational database. Each data table records spatiotemporal data at a specific subdivision level. Records at higher levels preserve pointers to the data table at lower levels. The GeoSOT-3D grid code is set as the primary key for each table. Temporal information is also recorded corresponding to each GeoSOT-3D code.</p> Full article ">Figure 4
<p>Collision detection for a single grid. A single spatiotemporal grid with Code<sub>A</sub> and a time interval from T<sub>A</sub> to T<sub>A</sub>’ is compared with the existing data at multiple levels. After determining the spatial relationship, the spatiotemporal grid information can be inserted into multiple data tables in parallel.</p> Full article ">Figure 5
<p>Collision detection for multiple grids. Multiple spatiotemporal grids are compared individually with the existing data at multiple levels. After determining the spatial relationship, the spatiotemporal grid information can be inserted into multiple data tables in parallel.</p> Full article ">Figure 6
<p>Time required for detection with different GeoSOT-3D grid sizes. The three selected GeoSOT-3D grid sizes are 128 m, 64 m, and 32 m, which correspond to levels 19, 20, and 21, respectively.</p> Full article ">Figure 7
<p>Time required for detecting collisions with three methods. The three methods consist of calculations with three-dimensional coordinates, calculations with simply 64 m grids, and spatial database queries with 64 m GeoSOT-3D grids.</p> Full article ">Figure 8
<p>Time required to detect collisions with three methods in an airspace containing obstacles. In simulation experiment (<b>a</b>), a 1 km × 1 km × 1 km obstacle was placed at the center of the airspace. In simulation experiment (<b>b</b>), 100 obstacles of the same size as the UAVs were placed within the airspace.</p> Full article ">Figure 9
<p>Time required to detect a collision after formation optimization. Collision detection with optimization is shown with a green line. The optimization is based on the premise that the UAVs are established in a fixed formation.</p> Full article ">
17 pages, 15310 KiB  
Article
City Maker: Reconstruction of Cities from OpenStreetMap Data for Environmental Visualization and Simulations
by I. Alihan Hadimlioglu and Scott A. King
ISPRS Int. J. Geo-Inf. 2019, 8(7), 298; https://doi.org/10.3390/ijgi8070298 - 15 Jul 2019
Cited by 15 | Viewed by 6884
Abstract
Recent innovations in 3D processing and availability of geospatial data have contributed largely to more comprehensive solutions to data visualization. As various data formats are utilized to describe the data, a combination of layers from different sources allow us to represent 3D urban [...] Read more.
Recent innovations in 3D processing and availability of geospatial data have contributed largely to more comprehensive solutions to data visualization. As various data formats are utilized to describe the data, a combination of layers from different sources allow us to represent 3D urban areas, contributing to ideas of emergency management and smart cities. This work focuses on 3D urban environment reconstruction using crowdsourced OpenStreetMap data. Once the data are extracted, the visualization pipeline draws features using coloring for added context. Moreover, by structuring the layers and entities through the addition of simulation parameters, the generated environment is made simulation ready for further use. Results show that urban areas can be properly visualized in 3D using OpenStreetMap data given data availability. The simulation-ready environment was tested using hypothetical flooding scenarios, which demonstrated that the added parameters can be utilized in environmental simulations. Furthermore, an efficient restructuring of data was implemented for viewing the city information once the data are parsed. Full article
►▼ Show Figures

Figure 1

Figure 1
<p>Different phases of City Maker.</p> Full article ">Figure 2
<p>Parts of feature extraction and drawing hierarchy of City Maker.</p> Full article ">Figure 3
<p>Processes City Maker utilizes to produce a city model.</p> Full article ">Figure 4
<p>A Multipolygon construct defined by relations, ways, nodes and tags.</p> Full article ">Figure 5
<p>Features currently implemented for the City Maker.</p> Full article ">Figure 6
<p>Visualization of Ward Island, home to Texas A&amp;M University—Corpus Christi, and parts of Ocean Drive.</p> Full article ">Figure 7
<p>3D shaded visualization of parts of Corpus Christi.</p> Full article ">Figure 8
<p>Visualization of six different cities: (<b>a</b>) Corpus Christi, TX, USA; (<b>b</b>) Frankfurt, Germany; (<b>c</b>) Izmit, Turkey; (<b>d</b>) Marseille, France; (<b>e</b>) San Jose, CA, USA; and (<b>f</b>) Zagreb, Croatia.</p> Full article ">Figure 9
<p>Parts of the city of Corpus Christi with storm sewer inlets.</p> Full article ">Figure 10
<p>Storm sewer information loaded in the City Maker.</p> Full article ">Figure 11
<p>Visualization of the depth in 2D for 50,000 particles: (<b>a</b>) with no friction coefficients; and (<b>b</b>) with friction coefficients.</p> Full article ">Figure 12
<p>Visualization of the flood simulation environment for 130,000 particles: (<b>a</b>) with no absorption coefficients; and (<b>b</b>) with absorption coefficients.</p> Full article ">
23 pages, 3581 KiB  
Article
A GIS-Based Support Vector Machine Model for Flash Flood Vulnerability Assessment and Mapping in China
by Junnan Xiong, Jin Li, Weiming Cheng, Nan Wang and Liang Guo
ISPRS Int. J. Geo-Inf. 2019, 8(7), 297; https://doi.org/10.3390/ijgi8070297 - 12 Jul 2019
Cited by 66 | Viewed by 6755
Abstract
Flash floods are one of the natural disasters that threaten the lives of many people all over the world every year. Flash floods are significantly affected by the intensification of extreme climate events and interactions with exposed and vulnerable socio-economic systems impede regional [...] Read more.
Flash floods are one of the natural disasters that threaten the lives of many people all over the world every year. Flash floods are significantly affected by the intensification of extreme climate events and interactions with exposed and vulnerable socio-economic systems impede regional development processes. Hence, it is important to estimate the loss due to flash floods before the disaster occurs. However, there are no comprehensive vulnerability assessment results for flash floods in China. Fortunately, the National Mountain Flood Disaster Investigation Project provided a foundation to develop this proposed assessment. In this study, an index system was established from the exposure and disaster reduction capability categories, and is based on analytic hierarchy process (AHP) methods. We evaluated flash flood vulnerability by adopting the support vector machine (SVM) model. Our results showed 439 counties with high and extremely high vulnerability (accounting for 10.5% of the land area and corresponding to approximately 100 million hectares (ha)), 571 counties with moderate vulnerability (accounting for 19.18% of the land area and corresponding to approximately 180 million ha), and 1128 counties with low and extremely low vulnerability (accounting for 39.43% of the land area and corresponding to approximately 370 million ha). The highly-vulnerable counties were mainly concentrated in the south and southeast regions of China, moderately-vulnerable counties were primarily concentrated in the central, northern, and southwestern regions of China, and low-vulnerability counties chiefly occurred in the northwest regions of China. Additionally, the results of the spatial autocorrelation suggested that the “High-High” values of spatial agglomeration areas mainly occurred in the Zhejiang, Fujian, Jiangxi, Hunan, Guangxi, Chongqing, and Beijing areas. On the basis of these results, our study can be used as a proposal for population and building distribution readjustments, and the management of flash floods in China. Full article
►▼ Show Figures

Figure 1

Figure 1
<p>The digital elevation model (DEM) of China.</p> Full article ">Figure 2
<p>Flowchart of vulnerability for flash floods in China.</p> Full article ">Figure 3
<p>Flow chart of quantification process of flash flood vulnerability index.</p> Full article ">Figure 4
<p>Spatial distribution pattern of exposure in China.</p> Full article ">Figure 5
<p>Spatial pattern of disaster reduction capabilities in China (the areas shown in the red ellipse are scored high).</p> Full article ">Figure 6
<p>Spatial pattern of (<b>a</b>) vulnerabilities and (<b>b</b>) local indicators of spatial association (LISAs) in China.</p> Full article ">Figure 7
<p>Moran’s <span class="html-italic">I</span> scatter plot of (<b>a</b>) exposure, (<b>b</b>) disaster reduction, and (<b>c</b>) vulnerability in counties of China.</p> Full article ">Figure 8
<p>Vulnerability in different geomorphological regions of China.</p> Full article ">
24 pages, 4108 KiB  
Review
Application of Remote Sensing to the Investigation of Rock Slopes: Experience Gained and Lessons Learned
by Doug Stead, Davide Donati, Andrea Wolter and Matthieu Sturzenegger
ISPRS Int. J. Geo-Inf. 2019, 8(7), 296; https://doi.org/10.3390/ijgi8070296 - 27 Jun 2019
Cited by 34 | Viewed by 7116
Abstract
The stability and deformation behavior of high rock slopes depends on many factors, including geological structures, lithology, geomorphic processes, stress distribution, and groundwater regime. A comprehensive mapping program is, therefore, required to investigate and assess the stability of high rock slopes. However, slope [...] Read more.
The stability and deformation behavior of high rock slopes depends on many factors, including geological structures, lithology, geomorphic processes, stress distribution, and groundwater regime. A comprehensive mapping program is, therefore, required to investigate and assess the stability of high rock slopes. However, slope steepness, rockfalls and ongoing instability, difficult terrain, and other safety concerns may prevent the collection of data by means of traditional field techniques. Therefore, remote sensing methods are often critical to perform an effective investigation. In this paper, we describe the application of field and remote sensing approaches for the characterization of rock slopes at various scale and distances. Based on over 15 years of the experience gained by the Engineering Geology and Resource Geotechnics Research Group at Simon Fraser University (Vancouver, Canada), we provide a summary of the potential applications, advantages, and limitations of varied remote sensing techniques for comprehensive characterization of rock slopes. We illustrate how remote sensing methods have been critical in performing rock slope investigations. However, we observe that traditional field methods still remain indispensable to collect important intact rock and discontinuity condition data. Full article
(This article belongs to the Special Issue Applications of Photogrammetry for Environmental Research)
►▼ Show Figures

Figure 1

Figure 1
<p>Location of the sites described in this paper. <b>1</b>: Hope Slide (British Columbia); <b>2</b>: Block 731 (British Columbia); <b>3</b>: Palliser rockslide (Alberta); <b>4</b>: Frank Slide (Alberta); <b>5</b>: Vajont Slide (Italy).</p> Full article ">Figure 2
<p>Remote sensing equipment employed at the investigated sites. <b>a</b>: Canon EOS 5Ds-R with <span class="html-italic">f</span> = 400 telephoto lens; <b>b</b>: DJI Phantom 3 Pro Quadcopter; <b>c</b>: Optech ILRIS3D terrestrial laser scanning (TLS); <b>d</b>: Riegl VZ-4000 TLS; <b>e</b>: FLIR SC7750 thermal camera; <b>f</b>: Specim SWIR3 hyperspectral scanner.</p> Full article ">Figure 3
<p>Remote sensing analysis performed at the Hope Slide. <b>a</b>: pre-failure 3D model of the slope, constructed using 1961 aerial imagery processed using the structure-from-motion (SfM) technique; <b>b</b>: example of a 3D model reconstructed using terrestrial digital photogrammetry (TDP). Blue and pink disks show discontinuities mapped in 3DM Analyst; <b>c</b>: structural overview of the area. The stereonets show the orientation of the discontinuity sets mapped in the TDP models, subdivided into five structural domains; <b>d</b>: orthorectified image of the debris field reconstructed using unmanned aerial vehicle (UAV)-SfM; <b>e</b>: infrared thermography (IRT) dataset of the daylighting sliding surface. Dark areas identify seepage.</p> Full article ">Figure 4
<p>Multi-sensor remote sensing analysis of the Block 731 road cut on the left abutment of the Revelstoke Dam. <b>a</b>: Panoramic overview of the 55 m high rock slope. Note the location and orientation of the shear zones S3 and S4; <b>b</b>: point cloud of the rock slope. The yellow box in <b>a</b> and <b>b</b> shows the location of sections in <b>c</b> and <b>d</b>; <b>c</b>: minimum noise fraction (MNF) image of part of the lower slope comprising the three lowest noise fraction bands obtained from the hyperspectral imagery (HSI) dataset. Details of augens and lithological alternations are shown. Note the high contrast between different geological materials; <b>d</b>: high-resolution photograph of the same part of the slope shown in <b>c</b>.</p> Full article ">Figure 5
<p>Remote sensing analysis performed at the Palliser Rockslide. <b>a:</b> panoramic photograph of the slope; <b>b</b>: TLS 3D point cloud; <b>c</b>: high-resolution photograph showing evidence of intra-bedding intact rock fracturing; <b>d</b>–<b>f</b>: details from the high-resolution photography, TLS, and IRT datasets showing areas of surface alteration (<b>a</b>1–<b>a</b>4) due to ephemeral seepage or concentrated water flow; <b>g</b>–<b>i</b>: details from the same datasets showing areas of surface alteration (<b>a</b>5) and karst features (<b>k</b>1, <b>k</b>2).</p> Full article ">Figure 6
<p>Multi-sensor remote sensing imagery of the Frank Slide headscarp. <b>a</b>: panoramic overview; <b>b</b>: point cloud of the South Peak area. Red and yellow boxes outline the sections shown in <b>b</b>–<b>e</b>, respectively; <b>c</b>: High-resolution photograph draped onto the TLS dataset; <b>d</b>: IRT image of the Frank Slide. Low temperature areas are associated to surface topography variation; <b>e</b>: HSI image of the Frank Slide (Red band: 1047.0 nm; green band: 1601.9 nm; blue band: 2103.2 nm). Variations in irradiance show possible mineralogical changes between layers.</p> Full article ">Figure 7
<p>Overview of the remote sensing analysis undertaken at the Vajont Slide site. <b>a</b>: geomorphological analysis of the pre-1963 slope, based on historic aerial photo-interpretation; <b>b</b>: oblique image of the failed slope and slide deposit. The boxes shows outline the location of the 3D TDP models shown in <b>c</b>–<b>e</b>; <b>c</b>: lower hemisphere stereonet plot showing the discontinuity planes mapped across the rupture surface; <b>d</b>: 3D TDP model of part of the eastern basal rupture surface, constructed using <b>a</b> <span class="html-italic">f</span> = 200 mm focal lens length; <b>e</b>: 3D TDP model of the Massalezza Creek area, constructed using a <span class="html-italic">f</span> = 400 mm focal lens length (<b>a</b>,<b>c</b>–<b>e</b> are modified from [<a href="#B79-ijgi-08-00296" class="html-bibr">79</a>], by permission).</p> Full article ">Figure 8
<p>Workflow describing an approach for the comprehensive analysis of rock slopes.</p> Full article ">
15 pages, 2917 KiB  
Article
Predicting the Upcoming Services of Vacant Taxis near Fixed Locations Using Taxi Trajectories
by Chunchun Hu and Jean-Claude Thill
ISPRS Int. J. Geo-Inf. 2019, 8(7), 295; https://doi.org/10.3390/ijgi8070295 - 27 Jun 2019
Cited by 4 | Viewed by 3401
Abstract
Emerging on-line reservation services and special car services have greatly affected the development of the taxi industry. Surprisingly, taking a taxi is still a significant problem in many large cities. In this paper, we present an effective solution based on the Hidden Markov [...] Read more.
Emerging on-line reservation services and special car services have greatly affected the development of the taxi industry. Surprisingly, taking a taxi is still a significant problem in many large cities. In this paper, we present an effective solution based on the Hidden Markov Model to predict the upcoming services of vacant taxis that appear at some fixed locations and at specific times. The model introduces a weighted confusion matrix and a modified Viterbi algorithm, combining the factors of time of day and traffic conditions. In our framework, the hotspot or hidden states extraction is implemented through kernel density estimation (KDE) and fuzzy partitioning of traffic zones is done via a Fuzzy C Means (FCM) algorithm. We implement the proposed model on a large-scale dataset of taxi trajectories in Beijing. In this use case, tests demonstrate the high accuracy of the modeling framework in predicting the upcoming services of vacant taxis. We further analyze the factors affecting the predictive accuracy via a prediction accuracy analysis and prediction location evaluation. The findings of this paper can provide intelligence for the improvement of taxi services, to increase the passenger capacity of taxis and also to improve the probability of passengers finding taxis. Full article
►▼ Show Figures

Figure 1

Figure 1
<p>The framework of the predictive model.</p> Full article ">Figure 2
<p>An example of predictive modeling based on the hidden Markov model (HMM).</p> Full article ">Figure 3
<p>Kernel density estimation (KDE) map of taxi drop-off locations within five days in Beijing.</p> Full article ">Figure 4
<p>Prediction accuracy at 11 fixed locations: (<b>a</b>) During peak times; (<b>b</b>) During off-peak times.</p> Full article ">Figure 5
<p>Predictive accuracy ratio between peak and off-peak times.</p> Full article ">Figure 6
<p>The number of drop-offs during different time intervals within one day.</p> Full article ">Figure 7
<p>Prediction results compared with the observed trajectories using the test data.</p> Full article ">Figure 8
<p>Statistics of the deviations between the predicted and observed locations within each range.</p> Full article ">
34 pages, 8019 KiB  
Article
Automatic Discovery of Railway Train Driving Modes Using Unsupervised Deep Learning
by Han Zheng, Zanyang Cui and Xingchen Zhang
ISPRS Int. J. Geo-Inf. 2019, 8(7), 294; https://doi.org/10.3390/ijgi8070294 - 27 Jun 2019
Cited by 3 | Viewed by 3085
Abstract
Driving modes play vital roles in understanding the stochastic nature of a railway system and can support studies of automatic driving and capacity utilization optimization. Integrated trajectory data containing information such as GPS trajectories and gear changes can be good proxies in the [...] Read more.
Driving modes play vital roles in understanding the stochastic nature of a railway system and can support studies of automatic driving and capacity utilization optimization. Integrated trajectory data containing information such as GPS trajectories and gear changes can be good proxies in the study of driving modes. However, in the absence of labeled data, discovering driving modes is challenging. In this paper, instead of classical models (railway-specified feature extraction and classical clustering), we used five deep unsupervised learning models to overcome this difficulty. In these models, adversarial autoencoders and stacked autoencoders are used as feature extractors, along with generative adversarial network-based and Kullback–Leibler (KL) divergence-based networks as clustering models. An experiment based on real and artificial datasets showed the following: (i) The proposed deep learning models outperform the classical models by 27.64% on average. (ii) Integrated trajectory data can improve the accuracy of unsupervised learning by approximately 13.78%. (iii) The different performance rankings of models based on indices with labeled data and indices without labeled data demonstrate the insufficiency of people’s understanding of the existing modes. This study also analyzes the relationship between the discovered modes and railway carrying capacity. Full article
►▼ Show Figures

Figure 1

Figure 1
<p>Examples of modes of driving railway trains (MDRTs) from the perspective of speed-location profile. Each line represents a mode in this figure. We use a method based on phase [<a href="#B6-ijgi-08-00294" class="html-bibr">6</a>] to illustrate each example mode. Mode 0, the expected profile, has the phase structure A-Cr-B-Cr-B. Mode 1 has the phase structure A-Cr-B-A-Cr-B. Mode 2 has the phase structure A-Cr-Co. Mode 3 has the phase structure A-B-A-Cr. Mode 4 has the phase structure A-B-Cr-A. Mode 5 has the phase structure A-Cr-B-A-B, where ‘A’ represents the acceleration phase, ‘Cr’ represents the cruising phase (maintaining constant speed with throttle manipulation), ‘Co’ represents the coasting phase (train operation while the throttle is idle before braking), and ‘B’ represents the braking phase (for a station stop, for speed restriction or for a restrictive signal).</p> Full article ">Figure 2
<p>Built-in locomotive control and information systems and integrated trajectory data. LAIS refers to the train operation status information system, ATIS refers to the automatic railway car number identification system, and TAX refers to a comprehensive locomotive safety information monitoring device. All devices and sensors are arranged around the control host, which plays a role as a control center. The solid line represents a speed-location profile example. The dotted line represents a time-location profile example. The dashed line represents a gear-location profile example. The colored line represents a signal-location profile example. The line-dashed line represents a pressure-location profile example.</p> Full article ">Figure 3
<p>Framework of the proposed approach. The figure is roughly divided into three parts from top to bottom. The first part (blue) represents the basic parts of this methodology, including preprocessing, the unsupervised deep learning models, and the evaluation metric; the second part (black) is the specific details of each methodology part, where the flow is indicated by the arrows; and the third part (orange) illustrates the outputs of each part.</p> Full article ">Figure 4
<p>Outlying points in GPS trajectories (of integrated trajectory data). The latitudes and longitudes in the figure represent the spatial extent of the example trajectory data.</p> Full article ">Figure 5
<p>Network structure of the adversarial autoencoder clustering model (AAEC): The top adversarial network imposes a categorical distribution on the cluster representation, and the bottom adversarial network imposes a Gaussian distribution on the feature representation.</p> Full article ">Figure 6
<p>Network structure of deep embedded clustering (DEC). The first row is a stacked autoencoder that is initialized layer by layer, where each layer is a denoising autoencoder trained to reconstruct the previous layer’s output after corruption [<a href="#B40-ijgi-08-00294" class="html-bibr">40</a>]. The second row is the cluster layer, whose objective is minimizing the Kullback–Leibler (KL) divergence.</p> Full article ">Figure 7
<p>Network structure of AAE and KLD-based cluster (AAEKC). The first row is the cluster layer, whose objective is minimizing the Kullback–Leibler divergence. The remainder of this network is an adversarial autoencoder [<a href="#B42-ijgi-08-00294" class="html-bibr">42</a>], where the second row is a standard autoencoder that reconstructs data <math display="inline"><semantics> <mi>x</mi> </semantics></math> from a latent feature <math display="inline"><semantics> <mi>z</mi> </semantics></math>. The bottom row diagrams a second network trained to discriminatively predict whether a sample arises from the feature of the autoencoder or from a Gaussian distribution.</p> Full article ">Figure 8
<p>Network Structure of SAE and CatGAN (SAECC). The network consists of the Generator, Discriminator, Encoder, and Decoder. Among them, the Encoder and Decoder form the SAE, and the remainder constitutes the CatGAN.</p> Full article ">Figure 9
<p>Network structure of AAE and CatGAN (AAECC).</p> Full article ">Figure 10
<p>Cluster number determination by indices without ground-truth labels.</p> Full article ">Figure 11
<p>Illustration of modes in RD9. Figure (<b>a</b>–<b>m</b>) are images of the modes. Due to space limitations, we describe these modes in terms of speed-location profiles. Figure (<b>n</b>) is used to express the statistics of the data contained in each mode.</p> Full article ">Figure 12
<p>Illustration of evaluation indices of Step (ii).</p> Full article ">Figure 13
<p>Illustration of evaluation indices of Step (iii). The first row (from (<b>a</b>–<b>c</b>)) illustrates distributions of <math display="inline"><semantics> <mrow> <mi>D</mi> <mi>V</mi> <mi>I</mi> </mrow> </semantics></math>, <math display="inline"><semantics> <mrow> <mi>S</mi> <mi>P</mi> </mrow> </semantics></math>, and <math display="inline"><semantics> <mi>L</mi> </semantics></math>, and the second row (from (<b>d</b>–<b>e</b>)) illustrates those of <math display="inline"><semantics> <mrow> <mi>D</mi> <mi>B</mi> <mi>I</mi> </mrow> </semantics></math> and <math display="inline"><semantics> <mrow> <mi>C</mi> <mi>P</mi> </mrow> </semantics></math>.</p> Full article ">Figure 14
<p>Evaluation indices with ground-truth labels: (<b>a</b>) The indices of the five proposed models on the RD2 and PRD2 datasets. (<b>b</b>) The indices of the DEC model on the RD2 and PRD2 datasets. (<b>c</b>) The indices of the AAEC model on the RD2 and PRD2 datasets.</p> Full article ">
25 pages, 2487 KiB  
Review
Automatic (Tactile) Map Generation—A Systematic Literature Review
by Jakub Wabiński and Albina Mościcka
ISPRS Int. J. Geo-Inf. 2019, 8(7), 293; https://doi.org/10.3390/ijgi8070293 - 27 Jun 2019
Cited by 25 | Viewed by 6039
Abstract
This paper presents a systematic literature review that reflects the current state of research in the field of algorithms and models for map generalization, the existing solutions for automatic (tactile) map generation, as well as good practices for designing spatial databases for the [...] Read more.
This paper presents a systematic literature review that reflects the current state of research in the field of algorithms and models for map generalization, the existing solutions for automatic (tactile) map generation, as well as good practices for designing spatial databases for the purposes of automatic map development. A total number of over 500 primary studies were screened in order to identify the most relevant research on automatic (tactile) map generation from the last decade. The reviewed papers revealed many existing solutions in the field of automatic map production, as well as algorithms (e.g., Douglas–Peucker, Visvalingam–Whyatt) and models (e.g., GAEL, CartACom) for data generalization that might be used to transform traditional spatial data into the haptic form, suitable for blind and visually impaired people. However, it turns out that a comprehensive solution for automatic tactile map generation does not exist. Full article
►▼ Show Figures

Figure 1

Figure 1
<p>The course of the selection process together with the number of included and excluded papers in each step (source: own study).</p> Full article ">Figure 2
<p>Relation between publish year and number of papers (source: own study).</p> Full article ">Figure 3
<p>Number of primary studies providing answers to particular research questions (n—selected primary studies) (source: own study).</p> Full article ">Figure 4
<p>Weighted network visualization of associations between terms within identified primary studies (source: own study).</p> Full article ">Figure 5
<p>Number of appearances of generalization operators across primary studies (source: own study).</p> Full article ">Figure 6
<p>Reasons for the exclusion of certain papers from the review (source: own study).</p> Full article ">
22 pages, 4611 KiB  
Article
The Role of African Emerging Space Agencies in Earth Observation Capacity Building for Facilitating the Implementation and Monitoring of the African Development Agenda: The Case of African Earth Observation Program
by Mahlatse Kganyago and Paidamwoyo Mhangara
ISPRS Int. J. Geo-Inf. 2019, 8(7), 292; https://doi.org/10.3390/ijgi8070292 - 27 Jun 2019
Cited by 19 | Viewed by 5432
Abstract
AU-Agenda 2063 was adopted at the 24th Ordinary Session of the African Heads of State and Government in 2015 as the blueprint for the future development of the continent. Built upon the continent’s past experiences, challenges, and successes, AU-Agenda 2063 comprehensively describes the [...] Read more.
AU-Agenda 2063 was adopted at the 24th Ordinary Session of the African Heads of State and Government in 2015 as the blueprint for the future development of the continent. Built upon the continent’s past experiences, challenges, and successes, AU-Agenda 2063 comprehensively describes the strategic path for Africa’s future development in the next 50 years. Thus, the monitoring of its implementation in various African states is critical for ensuring sustainable development and track progress. However, the higher cost of collecting data for accurately and reliably monitoring the implementation of Agenda 2063 may hinder the progress towards achieving these goals. Satellite Earth observation provides ample data, and thus has provided opportunities for the development of novel products and services with the potential to support implementation, monitoring and reporting for AU-Agenda 2063 development imperatives. However, it has been limitedly exploited in Africa, as evidenced by lower research outputs and investments. This calls for increased capacity building in the use of available EO data and products for various users including decision makers to advance national, regional and continental priorities. The use of such data products is often hampered by the capability to understand the products and thus their value for addressing socio-economic challenges. This paper discusses the potential of Earth observation capacity building for supporting the implementation, monitoring of, and reporting towards achieving AU-Agenda 2063 development imperatives. Specifically, this paper identifies existing capacity building resources, including the role of open and free Earth observation data, open-source software, and product dissemination platforms that can be leveraged for supporting national development, service delivery and the achievement of AU-Agenda 2063 targets. Furthermore, the paper recognizes the importance of bilateral and multilateral partnerships in leveraging existing know-how, technology and other resources for advancing strategic goals of African emerging space agencies and promoting sustainable development, with examples from South African National Space Agency (SANSA). Then, the challenges and opportunities for capacity building and the wide adoption of EO in Africa are discussed in the context of AU-Agenda 2063. The paper thus concludes that EO capacity building is essential to address the skills and data gaps and increase the use of EO-based solutions for decision making in various sectors, critical for achieving AU-A2063. Full article
►▼ Show Figures

Figure 1

Figure 1
<p>AU-Agenda 2063 First-Ten-Years Implementation Plan (FTYIP) priority areas and goals (first and second rings, respectively) that can be directly addressed by common Earth observation applications (outer ring).</p> Full article ">Figure 2
<p>Examples of satellite data catalogues: (<b>a</b>) USGS Earth Explorer: <a href="https://earthexplorer.usgs.gov/" target="_blank">https://earthexplorer.usgs.gov/</a> containing data from Landsat 1–8, Sentinel 2, ResourceSat 1–2, and a variety of AVHRR and MODIS data products; (<b>b</b>) SANSA catalogue: <a href="http://catalogue.sansa.org.za/search/" target="_blank">http://catalogue.sansa.org.za/search/</a> containing data from SumbandilaSat, SPOT 2–7, Landsat 1–8, CBERS 2–4; (<b>c</b>) ESA Copernicus Open Access Hub: <a href="https://scihub.copernicus.eu/dhus/#/home" target="_blank">https://scihub.copernicus.eu/dhus/#/home</a> containing data from Sentinel 1–3; and (<b>d</b>) INPE catalogue: <a href="http://www.dgi.inpe.br/catalogo/" target="_blank">http://www.dgi.inpe.br/catalogo/</a> containing data from MODIS, Landsat 1–8, CBERS-2–4, ResourceSat 1 and 2.</p> Full article ">Figure 3
<p>Some of the popular open-source software useful for capacity building in EO using data from a variety of satellite missions as well as other Geospatial data.</p> Full article ">Figure 4
<p>Examples of EO-based food security and agricultural monitoring information products provided by various operating systems. (<b>a</b>) Vegetation condition index (VCI) indicating current vegetation health compared to long-term historical average based on METOP-AVHRR data; (<b>b</b>) Crop condition map showing poor to exceptional conditions per major crop at continental scale provided by GEGLAM’s Crop Monitor for Early Warning bulletins based on remotely sensed data, ground observations and field reports; (<b>c</b>) Agro-climatic indicators (NDVI, Minimum Temperature, Maximum Temperature, Cumulative Precipitation, Evaporative Stress Index, Soil Moisture, Daily Precipitation, and Growing Degree days) overproduction areas in Free State South Africa provided by GEOGLAM’s Crop Monitor for AMIS bulletins; (<b>d</b>) eMODIS temporally smoothed NDVI provided by FEWS NET system every 10 days; (<b>e</b>) Crop warning matrix based on NDVI and Standardized Precipitation Index (SPI) provided by ASAP system.</p> Full article ">Figure 4 Cont.
<p>Examples of EO-based food security and agricultural monitoring information products provided by various operating systems. (<b>a</b>) Vegetation condition index (VCI) indicating current vegetation health compared to long-term historical average based on METOP-AVHRR data; (<b>b</b>) Crop condition map showing poor to exceptional conditions per major crop at continental scale provided by GEGLAM’s Crop Monitor for Early Warning bulletins based on remotely sensed data, ground observations and field reports; (<b>c</b>) Agro-climatic indicators (NDVI, Minimum Temperature, Maximum Temperature, Cumulative Precipitation, Evaporative Stress Index, Soil Moisture, Daily Precipitation, and Growing Degree days) overproduction areas in Free State South Africa provided by GEOGLAM’s Crop Monitor for AMIS bulletins; (<b>d</b>) eMODIS temporally smoothed NDVI provided by FEWS NET system every 10 days; (<b>e</b>) Crop warning matrix based on NDVI and Standardized Precipitation Index (SPI) provided by ASAP system.</p> Full article ">Figure 5
<p>EO-based agricultural monitoring products developed at SANSA-based Landsat 8 OLI for national planted area monitoring.</p> Full article ">Figure 6
<p>Global Surface Water Explorer (<a href="https://global-surface-water.appspot.com/" target="_blank">https://global-surface-water.appspot.com/</a>) provides data on water occurrence (1984 to 2015), water seasonally (2014–2015), Annual Water Recurrence (1984–2015) and Water Occurrence Change Intensity (1984–2015).</p> Full article ">Figure 7
<p>Water-related information for supporting integrated water resources management in South Africa. (<b>a</b>) The National Seasonal Water Bodies Layer and (<b>b</b>) irrigated and non-irrigated fields for compliance monitoring to National Water Act No. 36 of 1998 and associated regulations.</p> Full article ">Figure 8
<p>Participants of the TIGER training workshop in Pretoria, South Africa held from 7–11 September 2015. The participants consisted of 30 water management professionals from 10 African countries, viz. Swaziland, Egypt, Ghana, Kenya, Democratic Republic of Congo, Uganda, Ethiopia, Morocco, Madagascar and South Africa.</p> Full article ">
21 pages, 20916 KiB  
Article
Evaluating the Suitability of Urban Expansion Based on the Logic Minimum Cumulative Resistance Model: A Case Study from Leshan, China
by Haijun Wang, Peihao Peng, Xiangdong Kong, Tingbin Zhang and Guihua Yi
ISPRS Int. J. Geo-Inf. 2019, 8(7), 291; https://doi.org/10.3390/ijgi8070291 - 26 Jun 2019
Cited by 7 | Viewed by 3537
Abstract
This paper focuses on the suitability of urban expansion in mountain areas against the background of accelerated urban development. Urbanization is accompanied by conflict and intense transformations of various landscapes, and is accompanied by social, economic, and ecological impacts. Evaluating the suitability of [...] Read more.
This paper focuses on the suitability of urban expansion in mountain areas against the background of accelerated urban development. Urbanization is accompanied by conflict and intense transformations of various landscapes, and is accompanied by social, economic, and ecological impacts. Evaluating the suitability of urban expansion (UE) and determining an appropriate scale is vital to solving urban environmental issues and realizing sustainable urban development. In mountain areas, the natural and social environments are different from those in the plains; the former is characterized by fragile ecology and proneness to geological disasters. Therefore, when evaluating the expansion of a mountain city, more factors need to be considered. Moreover, we need to follow the principle of harmony between nature and society according to the characteristics of mountain cities. Thus, when we evaluate the expansion of a mountain city, the key procedure is to establish a scientific evaluation system and explore the relationship between each evaluation factor and the urban expansion process. Taking Leshan (LS), China—a typical mountain city in the upper Yangtze River which has undergone rapid growth—as a case study, the logic minimum cumulative resistance (LMCR) model was applied to evaluate the suitability of UE and to simulate its direction and scale. The results revealed that: An evaluation system of resistance factors (ESRFs) was established according to the principle of natural and social harmony; the logic resistance surface (LRS) scientifically integrated multiple resistance factors based on the ESRF and a logic regression analysis. LRS objectively and effectively reflected the contribution and impact of each resistance factor to urban expansion. We found that landscape, geological hazards and GDP have had a great impact on urban expansion in LS. The expansion space of the mountain city is limited; the area of suitable expansion is only 23.5%, while the area which is unsuitable for expansion is 39.3%. In addition, it was found that setting up ecological barriers is an effective way to control unreasonable urban expansion in mountain cities. There is an obvious scale (grid size) effect in the evaluation of urban expansion in mountain cities; an evaluation of the suitable scale yielded the result of 90 m × 90 m. On this scale, taking the central district as the center, the urban expansion process will extend to the neighboring towns of Mianzhu, Suji, Juzi and Mouzi. Urban expansion should be controlled in terms of scale, especially in mountain cities. The most suitable urban size of LS is 132 km2.This would allow for high connectivity of urban-rural areas with the occupation of relatively few green spaces. Full article
(This article belongs to the Special Issue Urban Growth Modelling in an Era of Live Data Sets for Dynamic Space/Time Analysis and Simulation)
►▼ Show Figures

Graphical abstract

Graphical abstract
Full article ">Figure 1
<p>Urban expansion from 1984 to 2019 in LS (Google Maps).</p> Full article ">Figure 2
<p>The geographical location of Leshan (LS) and its distribution in the Chengdu and Chongqing (C-Y) urban agglomeration.</p> Full article ">Figure 3
<p>Three-dimensional geographic environment of LS and the surrounding area.</p> Full article ">Figure 4
<p>True color image of LS taken in June 2018 with a multi-spectral sensor mounted on the ZY03 satellite.</p> Full article ">Figure 5
<p>Flow chart of data preprocessing.</p> Full article ">Figure 6
<p>Flow chart of the improved Minimum Cumulative Resistance (Logic minimum cumulative resistance) model.</p> Full article ">Figure 7
<p>The digital grid consists of the basic cell representing the LMCR value, and the cell with no data denotes the ecological barrier. The four digital grids (<b>A</b>, <b>B</b>, <b>C</b>, and <b>D</b>) represent the LMCR surface, suitability zone, the resistance variation in different directions and the calculation of expansion paths, respectively. Moreover, the urban geographic center (UGC) denotes the urban geographic center and <math display="inline"><semantics> <mi>t</mi> </semantics></math> indicates the satellite town.</p> Full article ">Figure 8
<p>Resistance curves of urban expansion. (<b>a</b>), (<b>b</b>), (<b>c</b>), and (<b>d</b>) are the resistance curves from UGC to Suji, Mianzhu, Juzi, and Mouzi, respectively.</p> Full article ">Figure 9
<p>Overlay layers of the LMCR surface, eco-barrier, and RF surfaces.</p> Full article ">Figure 10
<p>Directions and paths of urban expansion in LS.</p> Full article ">Figure 11
<p>Connectivity and scales in the different scenarios.</p> Full article ">Figure 12
<p>Urban planning map of LS from 2018 to 2030 (2017 version, LS government).</p> Full article ">
19 pages, 4004 KiB  
Article
High-Performance Overlay Analysis of Massive Geographic Polygons That Considers Shape Complexity in a Cloud Environment
by Kang Zhao, Baoxuan Jin, Hong Fan, Weiwei Song, Sunyu Zhou and Yuanyi Jiang
ISPRS Int. J. Geo-Inf. 2019, 8(7), 290; https://doi.org/10.3390/ijgi8070290 - 26 Jun 2019
Cited by 17 | Viewed by 5083
Abstract
Overlay analysis is a common task in geographic computing that is widely used in geographic information systems, computer graphics, and computer science. With the breakthroughs in Earth observation technologies, particularly the emergence of high-resolution satellite remote-sensing technology, geographic data have demonstrated explosive growth. [...] Read more.
Overlay analysis is a common task in geographic computing that is widely used in geographic information systems, computer graphics, and computer science. With the breakthroughs in Earth observation technologies, particularly the emergence of high-resolution satellite remote-sensing technology, geographic data have demonstrated explosive growth. The overlay analysis of massive and complex geographic data has become a computationally intensive task. Distributed parallel processing in a cloud environment provides an efficient solution to this problem. The cloud computing paradigm represented by Spark has become the standard for massive data processing in the industry and academia due to its large-scale and low-latency characteristics. The cloud computing paradigm has attracted further attention for the purpose of solving the overlay analysis of massive data. These studies mainly focus on how to implement parallel overlay analysis in a cloud computing paradigm but pay less attention to the impact of spatial data graphics complexity on parallel computing efficiency, especially the data skew caused by the difference in the graphic complexity. Geographic polygons often have complex graphical structures, such as many vertices, composite structures including holes and islands. When the Spark paradigm is used to solve the overlay analysis of massive geographic polygons, its calculation efficiency is closely related to factors such as data organization and algorithm design. Considering the influence of the shape complexity of polygons on the performance of overlay analysis, we design and implement a parallel processing algorithm based on the Spark paradigm in this paper. Based on the analysis of the shape complexity of polygons, the overlay analysis speed is improved via reasonable data partition, distributed spatial index, a minimum boundary rectangular filter and other optimization processes, and the high speed and parallel efficiency are maintained. Full article
(This article belongs to the Special Issue Big Data Computing for Geospatial Applications)
►▼ Show Figures

Figure 1

Figure 1
<p>Polygon overlay.</p> Full article ">Figure 2
<p>Diagram of azimuth interval calculation.</p> Full article ">Figure 3
<p>Hilbert partitioning and Hilbert curve generation.</p> Full article ">Figure 4
<p>Parallel overlay computing flow.</p> Full article ">Figure 5
<p>Extracting slope as the clipping layer from the digital elevation model (DEM).</p> Full article ">Figure 6
<p>Polygons distribution with different number of vertices.</p> Full article ">Figure 7
<p>Polygons distribution with different number of vertices.</p> Full article ">Figure 8
<p>Time consumption statistical graphs of different computing modes.</p> Full article ">Figure 9
<p>Time consumption comparison of different optimization strategies.</p> Full article ">Figure 10
<p>Average running time of different numbers of nodes.</p> Full article ">Figure 11
<p>Acceleration ratio of different numbers of nodes.</p> Full article ">Figure 12
<p>Parallel efficiency of different number of nodes.</p> Full article ">
Show export options expand_more Show export options expand_less
Select all
Export citation of selected articles as:
 
Displaying articles 1-26
Previous Issue
Next Issue
Back to TopTop