A Two-Step Integrated MLP-GTWR Method to Estimate 1 km Land Surface Temperature with Complete Spatial Coverage in Humid, Cloudy Regions
<p>Basic information about the study area of South China. (<b>a</b>) The land cover category in 2015 from the Moderate Resolution Imaging Spectroradiometer (MODIS) Land Cover Type MCD12Q1 version 6 data product (<a href="https://lpdaac.usgs.gov/products/mcd12q1v006/" target="_blank">https://lpdaac.usgs.gov/products/mcd12q1v006/</a> (accessed on 12 November 2020).) and the spatial distribution of four ChinaFLUX ground flux towers. (<b>b</b>) The percentage of valid clear-sky observations in the MODIS MYD11A1 land surface temperature (LST) product and the spatial locations of ten automatic meteorological stations. The slash or blue areas indicate the pixels with full of water surface or not belonging to the five Köppen–Geiger climate types in South China.</p> "> Figure 2
<p>Schematic illustration of the multilayer perceptron-geographically and temporally weighted regression (MLP-GTWR) method.</p> "> Figure 3
<p>Spatial patterns and frequency distribution curves of statistical metrics between the MLP-GTWR LST (1 km) and MODIS LST (1 km) for the pixels without being used in aggregation to the 10-km MODIS LST. (<b>a</b>) The spatial pattern of the valid observations (Days) of MODIS LST (1 km) without being used in aggregation to the 10-km MODIS LST. The higher pixel number in the narrow zone of the seashore is because the pixels have a large part of water surface and are not used in aggregation to the coarse-resolution LST. (<b>b</b>–<b>e</b>) The spatial patterns of the bias, MAE, RMSE, and R<sup>2</sup>. (<b>f</b>) The frequency distribution curves of above four statistical metrics.</p> "> Figure 4
<p>Scatterplots of the MODIS LST and the MLP-GTWR LST<sub>cor</sub> against the in situ LST at four flux stations with valid data number (N), MAE, RMSE, and R<sup>2</sup>. The left side is the MODIS LST, and the right is the MLP-GTWR LST<sub>cor</sub>. (<b>a</b>,<b>b</b>) Qianyanzhou, (<b>c</b>,<b>d</b>) Dinghushan, (<b>e</b>,<b>f</b>) Taoyuan, and (<b>g</b>,<b>h</b>) Xishuangbanna.</p> "> Figure 5
<p>Comparisons of the 1-km MODIS LST, aggregated 10-km MODIS LST, 10-km MLP LST, 1-km MLP-GTWR LST, and bias-corrected MLP-GTWR LST<sub>cor</sub> (1 km) against the in situ LST at four flux stations: Qianyanzhou (<b>a</b>), Dinghushan (<b>b</b>), Taoyuan (<b>c</b>), and Xishuangbanna (<b>d</b>), respectively.</p> "> Figure 6
<p>Spatial patterns of daytime 1-km MODIS LST (the first row), 10-km MLP LST (the second row), 1-km MLP-GTWR LST<sub>cor</sub> (the third row), and 10-km ERA5-Land LST (the fourth row) on the 14th (the first column), 105th (the second column), 197th (the third column), and 289th (the fourth column) days in 2014. The blank areas indicate the missing MODIS LST due primarily to cloud contamination and the AMSR2 gaps due to its orbit swath.</p> "> Figure 7
<p>Spatial patterns of nighttime 1-km MODIS LST (the first row), 10-km MLP LST (the second row), 1-km MLP-GTWR LSTcor (the third row), and 10-km ERA5-Land LST (the fourth row) on the 14th (the first column), 105th (the second column), 197th (the third column), and 289th (the fourth column) days in 2014. The blank areas indicate the missing MODIS LST due primarily to cloud contami-nation and the AMSR2 gaps due to its orbit swath.</p> "> Figure 8
<p>Temporal variations in daytime and nighttime MODIS LST, MLP-GTWR LST<sub>cor</sub>, and in situ measurements at four flux stations: Qianyanzhou (<b>a</b>), Dinghushan (<b>b</b>), Taoyuan (<b>c</b>), and Xishuangbanna (<b>d</b>). The subplots are seasonal statistics of MLP-GTWR LST<sub>cor</sub> against in situ measurements. The square dots give the MAE for daytime and nighttime, with the length of the line centered at each square dot representing one standard deviation of the difference.</p> "> Figure 9
<p>The relative contributions of predictors to the increased RMSE by an important test in the MLP process. AMSR 6.9 (H/V) means the 6.9 GHz channel of the AMSR2 BTs in horizontal (H) or vertical (V) polarization, and the others are similar; DEM: digital elevation model; FVC: fraction of vegetation coverage; LAI: leaf area index.</p> "> Figure 10
<p>The relative contributions of four land surfaces to the LST downscaling in the GTWR process presented by the Pratt index (transformed to percentage). (<b>a</b>) FVC: fraction of vegetation coverage; (<b>b</b>) LAI: leaf area index; (<b>c</b>) DEM: digital elevation model; (<b>d</b>) NL: nighttime lights.</p> "> Figure 11
<p>Comparison of the MLP-GTWR LST<sub>cor</sub> and two reconstructed LST using the method in Li et al. (2018) and Shwetha and Kumar (2016) against the in situ measurements.</p> "> Figure A1
<p>High-resolution satellite imagery (Google Earth) of the surrounding areas of the four flux stations listed in Tab. 1. The position of the stations is in the center and marked by the black-circle dots. (<b>a</b>,<b>b</b>) Qianyanzhou, (<b>c</b>,<b>d</b>) Dinghushan, (<b>e</b>,<b>f</b>) Taoyuan, and (<b>g</b>,<b>h</b>) Xishuangbanna. The left and right column is at the scale of 10 km and 1 km, respectively. The red rectangles in the left column mark the areas in the right column. The scale legend and north arrow is located at the bottom right corner.</p> "> Figure A2
<p>Sensitivity analysis of the node numbers of the two-layer multilayer perceptron model and the length of sliding window. (<b>a</b>) Performance change with the first hidden layer with the fixed node 10 in the second hidden layer. (<b>b</b>) Performance change with the second hidden layer with the fixed node 14 in the first hidden layer. (<b>c</b>) Performance change with the window length with the fixed sliding distance 10 km (one grid’s size). The RMSE is calculated from the comparison of the test sets (15% of the training dataset) and their corresponding predictions in the MLP model. Considering the computing time cost, the final nodes in two layers are chosen to be 14 and 7, and the window size is 90 in the MLP process.</p> "> Figure A3
<p>The precipitation recorded by the ground meteorological stations at the 197th day in 2014 across South China. (<b>a</b>) the total precipitation in the period of 20:00-08:00. (<b>b</b>) the total precipitation in the period of 08:00-20:00. (<b>c</b>) the total precipitation of both (<b>a</b>) and (<b>b</b>). The precipitation data is downloaded from China Meteorological Administration ().</p> "> Figure A4
<p>Temporal variations in daytime and nighttime MODIS LST, MLP-GTWR LSTcor, and in-situ measurements at Dinghushan station in the year 2014 (<b>top</b>) and 2015 (<b>bottom</b>).</p> ">
Abstract
:1. Introduction
2. Study Area and Dataset
2.1. Study Area
2.2. Satellite-Based Datasets
2.2.1. Moderate Resolution Imaging Spectroradiometer (MODIS) Land Surface Temperature (LST) Data
2.2.2. Advanced Microwave Scanning Radiometer (AMSR2) Brightness Temperature Data
2.2.3. Land Surface Feature Data
2.3. In Situ Measurements
2.4. Other Validation Data
3. Methods
3.1. Multilayer Perceptron (MLP) Process with Moving Windows to Generate 10-km All-Weather Land Surface Temperature
- (1)
- Set a square window and create the training data with all available MODIS LST in this window.
- (2)
- Train an MLP model used only in this window.
- (3)
- Predict the 10-km LST map using the trained MLP model.
- (4)
- Slide the window with a distance in the direction of right or down, and repeat steps 1–4 until covering the entire study area.
3.2. Geographically and Temporally Weighted Regression (GTWR) Process to Downscale Land Surface Temperature
3.3. Bias Correction
3.4. Evaluation
4. Results
4.1. Performance Evaluation of the MLP-GTWR Method Evaluated with MODIS LST
4.2. Performance Evaluation of the MLP-GTWR Method Evaluated with In Situ Measurements
4.3. Spatial Variability in the MLP-GTWR LSTcor
4.4. Temporal Variability in the MLP-GTWR LSTcor
5. Discussion
5.1. The Advantages of the MLP and GTWR Processes in the Proposed Two-Step Method
5.2. The Relative Importance of the Predictors in the MLP-GTWR Method
5.3. Comparison of the MLP-GTWR Method with a Gap-Filling Method and a Same-Type Method
5.4. Limitations and Potential for Further Applications
6. Conclusions
Author Contributions
Funding
Institutional Review Board Statement
Informed Consent Statement
Data Availability Statement
Acknowledgments
Conflicts of Interest
Appendix A
Station | Total | Spring | Summer | Autumn | Winter |
---|---|---|---|---|---|
Qianyanzhou | 14 | 2 | 4 | 5 | 3 |
Dinghushan | 8 | 2 | 2 | 3 | 1 |
Taoyuan | 7 | 2 | 2 | 2 | 1 |
Xishuangbanna | 11 | 4 | 1 | 3 | 3 |
Metrics | Qianyanzhou | Dinghushan | Taoyuan | Xishuangbanna | ||||
---|---|---|---|---|---|---|---|---|
MLP-GTWR | Li’s Approach | MLP-GTWR | Li’s Approach | MLP-GTWR | Li’s Approach | MLP-GTWR | Li’s Approach | |
Bias | −0.42 | −0.85 | −1.99 | −1.83 | −0.94 | −0.54 | 2.16 | 2.16 |
MAE | 2.25 | 4.01 | 2.94 | 3.11 | 2.47 | 3.24 | 2.50 | 2.96 |
RMSE | 2.81 | 5.26 | 3.55 | 3.92 | 2.99 | 4.03 | 3.01 | 3.80 |
R2 | 0.91 | 0.75 | 0.87 | 0.82 | 0.95 | 0.85 | 0.94 | 0.84 |
Metrics | Qianyanzhou | Dinghushan | Taoyuan | Xishuangbanna | ||||
---|---|---|---|---|---|---|---|---|
MLP-GTWR | Shwetha ’s Approach | MLP-GTWR | Shwetha ’s Approach | MLP-GTWR | Shwetha ’s Approach | MLP-GTWR | Shwetha ’s Approach | |
Bias | −0.51 | 1.62 | −1.88 | −1.08 | −0.93 | 1.53 | 1.76 | 2.69 |
MAE | 2.35 | 4.33 | 2.90 | 3.56 | 2.41 | 3.47 | 2.25 | 3.33 |
RMSE | 2.95 | 5.68 | 3.51 | 4.33 | 2.91 | 4.64 | 2.81 | 4.42 |
R2 | 0.89 | 0.65 | 0.85 | 0.69 | 0.91 | 0.79 | 0.90 | 0.75 |
References
- Janatian, N.; Sadeghi, M.; Sanaeinejad, S.H.; Bakhshian, E.; Farid, A.; Hasheminia, S.M.; Ghazanfari, S. A statistical framework for estimating air temperature using MODIS land surface temperature data: Estimating air temperature using modis land surface temperature. Int. J. Climatol. 2017, 37, 1181–1194. [Google Scholar] [CrossRef]
- Silvestro, F.; Gabellani, S.; Delogu, F.; Rudari, R.; Boni, G. Exploiting remote sensing land surface temperature in distributed hydrological modelling: The example of the Continuum Model. Hydrol. Earth Syst. Sci. 2013, 17, 39–62. [Google Scholar] [CrossRef] [Green Version]
- Zhang, Y.; Liang, S. Impacts of land cover transitions on surface temperature in China based on satellite observations. Environ. Res. Lett. 2018, 13, 024010. [Google Scholar] [CrossRef]
- Tomlinson, C.J.; Chapman, L.; Thornes, J.E.; Baker, C. Remote sensing land surface temperature for meteorology and climatology: A review. Meteorol. Appl. 2011, 18, 296–306. [Google Scholar] [CrossRef] [Green Version]
- Srivastava, P.K.; Han, D.; Ramirez, M.R.; Islam, T. Machine learning techniques for downscaling SMOS satellite soil moisture using MODIS land surface temperature for hydrological application. Water Resour. Manag. 2013, 27, 3127–3144. [Google Scholar] [CrossRef]
- Colliander, A.; Fisher, J.B.; Halverson, G.; Merlin, O.; Misra, S.; Bindlish, R.; Jackson, T.J.; Yueh, S. Spatial downscaling of SMAP soil moisture using MODIS land surface temperature and NDVI during SMAPVEX15. IEEE Geosci. Remote Sens. Lett. 2017, 14, 2107–2111. [Google Scholar] [CrossRef]
- Hu, T.; Zhao, T.; Shi, J.; Wu, S.; Liu, D.; Qin, H.; Zhao, K. High-resolution mapping of freeze/thaw status in China via fusion of MODIS and AMSR2 data. Remote Sens. 2017, 9, 1339. [Google Scholar] [CrossRef] [Green Version]
- Quintano, C.; Fernandez-Manso, A.; Roberts, D.A. Burn severity mapping from landsat MESMA fraction images and land surface temperature. Remote Sens. Environ. 2017, 190, 83–95. [Google Scholar] [CrossRef]
- Khanal, S.; Fulton, J.; Shearer, S. An overview of current and potential applications of thermal remote sensing in precision agriculture. Comput. Electron. Agric. 2017, 139, 22–32. [Google Scholar] [CrossRef]
- Nguyen, H.; Wheeler, M.C.; Otkin, J.A.; Cowan, T.; Frost, A.; Stone, R. Using the evaporative stress index to monitor flash drought in Australia. Environ. Res. Lett. 2019, 14, 064016. [Google Scholar] [CrossRef] [Green Version]
- Phan, T.N.; Kappas, M. Application of MODIS land surface temperature data: A systematic literature review and analysis. J. Appl. Remote Sens. 2018, 12, 1. [Google Scholar] [CrossRef]
- Jin, M.; Dickinson, R.E. A generalized algorithm for retrieving cloudy sky skin temperature from satellite thermal infrared radiances. J. Geophys. Res. Atmospheres 2000, 105, 27037–27047. [Google Scholar] [CrossRef]
- Pede, T.; Mountrakis, G. An empirical comparison of interpolation methods for MODIS 8-day land surface temperature composites across the conterminous Unites States. ISPRS J. Photogramm. Remote Sens. 2018, 142, 137–150. [Google Scholar] [CrossRef]
- Yoo, C.; Im, J.; Cho, D.; Yokoya, N.; Xia, J.; Bechtel, B. Estimation of all-weather 1 Km MODIS land surface temperature for humid summer days. Remote Sens. 2020, 12, 1398. [Google Scholar] [CrossRef]
- Rao, Y.; Liang, S.; Wang, D.; Yu, Y.; Song, Z.; Zhou, Y.; Shen, M.; Xu, B. Estimating daily average surface air temperature using satellite land surface temperature and top-of-atmosphere radiation products over the tibetan plateau. Remote Sens. Environ. 2019, 234, 111462. [Google Scholar] [CrossRef]
- Li, Z.-L.; Tang, B.-H.; Wu, H.; Ren, H.; Yan, G.; Wan, Z.; Trigo, I.F.; Sobrino, J.A. Satellite-Derived Land Surface Temperature: Current Status and Perspectives. Remote Sens. Environ. 2013, 131, 14–37. [Google Scholar] [CrossRef] [Green Version]
- Ghafarian Malamiri, H.R.; Rousta, I.; Olafsson, H.; Zare, H.; Zhang, H. Gap-filling of MODIS time series land surface temperature (LST) products using singular spectrum analysis (SSA). Atmosphere 2018, 9, 334. [Google Scholar] [CrossRef] [Green Version]
- Kang, J.; Tan, J.; Jin, R.; Li, X.; Zhang, Y. Reconstruction of MODIS land surface temperature products based on multi-temporal information. Remote Sens. 2018, 10, 1112. [Google Scholar] [CrossRef] [Green Version]
- Li, X.; Zhou, Y.; Asrar, G.R.; Zhu, Z. Creating a seamless 1 Km resolution daily land surface temperature dataset for urban and surrounding areas in the conterminous United States. Remote Sens. Environ. 2018, 206, 84–97. [Google Scholar] [CrossRef]
- Yu, W.; Tan, J.; Ma, M.; Li, X.; She, X.; Song, Z. An effective similar-pixel reconstruction of the high-frequency cloud-covered areas of Southwest China. Remote Sens. 2019, 11, 336. [Google Scholar] [CrossRef] [Green Version]
- Wang, T.; Shi, J.; Ma, Y.; Husi, L.; Comyn-Platt, E.; Ji, D.; Zhao, T.; Xiong, C. Recovering land surface temperature under cloudy skies considering the solar-cloud-satellite geometry: Application to MODIS and landsat-8 data. J. Geophys. Res. Atmos. 2019, 124, 3401–3416. [Google Scholar] [CrossRef]
- Yang, Y.; Smith, J.; Yang, L.; Baeck, M.L.; Ni, G. Regional impacts of urban irrigation on surface heat fluxes and rainfall in Central Arizona. J. Geophys. Res. Atmos. 2019, 124, 6393–6410. [Google Scholar] [CrossRef]
- Zeng, C.; Long, D.; Shen, H.; Wu, P.; Cui, Y.; Hong, Y. A two-step framework for reconstructing remotely sensed land surface temperatures contaminated by cloud. ISPRS J. Photogramm. Remote Sens. 2018, 141, 30–45. [Google Scholar] [CrossRef]
- Fu, P.; Xie, Y.; Weng, Q.; Myint, S.; Meacham-Hensold, K.; Bernacchi, C. A Physical model-based method for retrieving urban land surface temperatures under cloudy conditions. Remote Sens. Environ. 2019, 230, 111191. [Google Scholar] [CrossRef]
- Long, D.; Yan, L.; Bai, L.; Zhang, C.; Li, X.; Lei, H.; Yang, H.; Tian, F.; Zeng, C.; Meng, X.; et al. Generation of MODIS-like land surface temperatures under all-weather conditions based on a data fusion approach. Remote Sens. Environ. 2020, 246, 111863. [Google Scholar] [CrossRef]
- Xia, H.; Chen, Y.; Li, Y.; Quan, J. Combining kernel-driven and fusion-based methods to generate daily high-spatial-resolution land surface temperatures. Remote Sens. Environ. 2019, 224, 259–274. [Google Scholar] [CrossRef]
- Rummukainen, M. State-of-the-art with regional climate models. WIREs Clim. Chang. 2010, 1, 82–96. [Google Scholar] [CrossRef]
- Duan, S.-B.; Li, Z.-L.; Leng, P. A Framework for the retrieval of all-weather land surface temperature at a high spatial resolution from polar-orbiting thermal infrared and passive microwave data. Remote Sens. Environ. 2017, 195, 107–117. [Google Scholar] [CrossRef]
- Huang, C.; Duan, S.-B.; Jiang, X.-G.; Han, X.-J.; Leng, P.; Gao, M.-F.; Li, Z.-L. A physically based algorithm for retrieving land surface temperature under cloudy conditions from AMSR2 passive microwave measurements. Int. J. Remote Sens. 2019, 40, 1828–1843. [Google Scholar] [CrossRef]
- Zhang, X.; Zhou, J.; Gottsche, F.-M.; Zhan, W.; Liu, S.; Cao, R. A Method based on temporal component decomposition for estimating 1-km all-weather land surface temperature by merging satellite thermal infrared and passive microwave observations. IEEE Trans. Geosci. Remote Sens. 2019, 57, 4670–4691. [Google Scholar] [CrossRef]
- Kou, X.; Jiang, L.; Bo, Y.; Yan, S.; Chai, L. Estimation of land surface temperature through blending MODIS and AMSR-E data with the bayesian maximum entropy method. Remote Sens. 2016, 8, 105. [Google Scholar] [CrossRef] [Green Version]
- Shwetha, H.R.; Kumar, D.N. Prediction of high spatio-temporal resolution land surface temperature under cloudy conditions using microwave vegetation index and ANN. ISPRS J. Photogramm. Remote Sens. 2016, 117, 40–55. [Google Scholar] [CrossRef]
- Holmes, T.R.H.; De Jeu, R.A.M.; Owe, M.; Dolman, A.J. Land surface temperature from Ka Band (37 GHz) passive microwave observations. J. Geophys. Res. 2009, 114, D04113. [Google Scholar] [CrossRef] [Green Version]
- Jiménez, C.; Prigent, C.; Ermida, S.L.; Moncet, J.-L. Inversion of AMSR-E observations for land surface temperature estimation: 1. methodology and evaluation with station temperature. J. Geophys. Res. Atmos. 2017, 122, 3330–3347. [Google Scholar] [CrossRef]
- Owe, M.; Griend, A.A.V.D. On the relationship between thermodynamic surface temperature and high-frequency (37 GHz) vertically polarized brightness temperature under semi-arid conditions. Int. J. Remote Sens. 2001, 22, 3521–3532. [Google Scholar] [CrossRef]
- Weng, F.; Grody, N.C. Physical retrieval of land surface temperature using the special sensor microwave imager. J. Geophys. Res. Atmos. 1998, 103, 8839–8848. [Google Scholar] [CrossRef]
- Chen, J.; Zhu, X.; Vogelmann, J.E.; Gao, F.; Jin, S. A simple and effective method for filling gaps in landsat ETM+ SLC-off images. Remote Sens. Environ. 2011, 115, 1053–1064. [Google Scholar] [CrossRef]
- Peng, Y.; Li, W.; Luo, X.; Li, H. A Geographically and temporally weighted regression model for spatial downscaling of MODIS land surface temperatures over urban heterogeneous regions. IEEE Trans. Geosci. Remote Sens. 2019, 57, 5012–5027. [Google Scholar] [CrossRef]
- Zhang, X.; Zhou, J.; Liang, S.; Chai, L.; Wang, D.; Liu, J. Estimation of 1-km all-weather remotely sensed land surface temperature based on reconstructed spatial-seamless satellite passive microwave brightness temperature and thermal infrared data. ISPRS J. Photogramm. Remote Sens. 2020, 167, 321–344. [Google Scholar] [CrossRef]
- Kottek, M.; Grieser, J.; Beck, C.; Rudolf, B.; Rubel, F. World map of the Köppen-Geiger climate classification updated. Meteorol. Z. 2006, 15, 259–263. [Google Scholar] [CrossRef]
- Wan, Z.; Dozier, J. A Generalized split-window algorithm for retrieving land-surface temperature from space. IEEE Trans. Geosci. Remote Sens. 1996, 34, 892–905. [Google Scholar] [CrossRef] [Green Version]
- Baret, F.; Weiss, M.; Lacaze, R.; Camacho, F.; Makhmara, H.; Pacholcyzk, P.; Smets, B. GEOV1: LAI and FAPAR essential climate variables and FCOVER global time series capitalizing over existing products. Part 1: Principles of development and production. Remote Sens. Environ. 2013, 137, 299–309. [Google Scholar] [CrossRef]
- Baugh, K.; Hsu, F.-C.; Elvidge, C.D.; Zhizhin, M. Nighttime lights compositing using the VIIRS day-night band: Preliminary results. Proc. Asia-Pac. Adv. Netw. 2013, 35, 70–86. [Google Scholar] [CrossRef] [Green Version]
- Duveiller, G.; Hooker, J.; Cescatti, A. The mark of vegetation change on earth’s surface energy balance. Nat. Commun. 2018, 9, 1–12. [Google Scholar] [CrossRef] [Green Version]
- Guo, W.; Lu, D.; Wu, Y.; Zhang, J. Mapping impervious surface distribution with integration of SNNP VIIRS-DNB and MODIS NDVI Data. Remote Sens. 2015, 7, 12459–12477. [Google Scholar] [CrossRef] [Green Version]
- Yu, G.-R.; Wen, X.-F.; Sun, X.-M.; Tanner, B.D.; Lee, X.; Chen, J.-Y. Overview of ChinaFLUX and evaluation of its eddy covariance measurement. Agric. For. Meteorol. 2006, 137, 125–137. [Google Scholar] [CrossRef]
- Ermida, S.L.; Soares, P.; Mantas, V.; Göttsche, F.-M.; Trigo, I.F. Google Earth Engine open-source code for land surface temperature estimation from the landsat series. Remote Sens. 2020, 12, 1471. [Google Scholar] [CrossRef]
- Duan, S.-B.; Li, Z.-L.; Li, H.; Göttsche, F.-M.; Wu, H.; Zhao, W.; Leng, P.; Zhang, X.; Coll, C. Validation of collection 6 MODIS land surface temperature product using in situ measurements. Remote Sens. Environ. 2019, 225, 16–29. [Google Scholar] [CrossRef] [Green Version]
- Göttsche, F.-M.; Olesen, F.-S.; Trigo, I.F.; Bork-Unkelbach, A.; Martin, M.A. Long term validation of land surface temperature retrieved from MSG/SEVIRI with continuous in-situ measurements in Africa. Remote Sens. 2016, 8, 410. [Google Scholar] [CrossRef] [Green Version]
- Wang, K. Estimation of surface long wave radiation and broadband emissivity using moderate resolution imaging spectroradiometer (modis) land surface temperature/emissivity products. J. Geophys. Res. 2005, 110, D11109. [Google Scholar] [CrossRef]
- Bergstra, J.; Bengio, Y. Random search for hyper-parameter optimization. J. Mach. Learn. Res. 2012, 13, 281–305. [Google Scholar]
- He, Q.; Huang, B. Satellite-based mapping of daily high-resolution ground PM2. 5 in China via space-time regression modeling. Remote Sens. Environ. 2018, 206, 72–83. [Google Scholar] [CrossRef]
- Huang, B.; Wu, B.; Barry, M. Geographically and temporally weighted regression for modeling spatio-temporal variation in house prices. Int. J. Geogr. Inf. Sci. 2010, 24, 383–401. [Google Scholar] [CrossRef]
- Brunsdon, C.; Fotheringham, S.; Charlton, M. Geographically weighted regression. J. R. Stat. Soc. Ser. Stat. 1998, 47, 431–443. [Google Scholar] [CrossRef]
- Fotheringham, A.S.; Brunsdon, C.; Charlton, M. Geographically Weighted Regression: The Analysis of Spatially Varying Relationships; John Wiley & Sons: Hoboken, NJ, USA, 2003. [Google Scholar]
- Chen, J.; Brissette, F.P.; Leconte, R. Uncertainty of downscaling method in quantifying the impact of climate change on hydrology. J. Hydrol. 2011, 401, 190–202. [Google Scholar] [CrossRef]
- Leander, R.; Buishand, T.A. Resampling of regional climate model output for the simulation of extreme river flows. J. Hydrol. 2007, 332, 487–496. [Google Scholar] [CrossRef]
- Li, X.; Zhang, L.; Weihermüller, L.; Jiang, L.; Vereecken, H. Measurement and simulation of topographic effects on passive microwave remote sensing over mountain areas: A case study from the Tibetan Plateau. IEEE Trans. Geosci. Remote Sens. 2014, 52, 1489–1501. [Google Scholar] [CrossRef]
- Smith, T.; Bookhagen, B. Assessing uncertainty and sensor biases in passive microwave data across high mountain Asia. Remote Sens. Environ. 2016, 181, 174–185. [Google Scholar] [CrossRef]
- McFarland, M.J.; Miller, R.L.; Neale, C.M.U. Land surface temperature derived from the SSM/I passive microwave brightness temperatures. IEEE Trans. Geosci. Remote Sens. 1990, 28, 839–845. [Google Scholar] [CrossRef]
- Forte, G.F.; Camps, A.; Tarongi, J.M.; Vall-Llossera, M. Study of radio frequency interference effects on radiometry bands in urban environments. In Proceedings of the 2012 IEEE International Geoscience and Remote Sensing Symposium, Munich, Germany, 22–27 July 2012; pp. 1069–1072. [Google Scholar]
- Dong, J.; Peng, J.; He, X.; Corcoran, J.; Qiu, S.; Wang, X. Heatwave-induced human health risk assessment in megacities based on heat stress-social vulnerability-human exposure framework. Landsc. Urban Plan. 2020, 203, 103907. [Google Scholar] [CrossRef]
- Yang, J.; Shen, K.; Ong, C.; Li, X. Feature selection for MLP neural network: The use of random permutation of probabilistic outputs. IEEE Trans. Neural Netw. 2009, 20, 1911–1922. [Google Scholar] [CrossRef]
- Chao, Y.-C.E.; Kupper, L.L.; Serdar, B.; Egeghy, P.P.; Rappaport, S.M.; Nylander-French, L.A. Dermal exposure to jet fuel JP-8 significantly contributes to the production of urinary naphthols in fuel-cell maintenance workers. Environ. Health Perspect. 2006, 114, 182–185. [Google Scholar] [CrossRef] [PubMed] [Green Version]
- Prigent, C.; Jimenez, C.; Aires, F. Toward “All Weather,” long record, and real-time land surface temperature retrievals from microwave satellite observations: Microwave land surface temperature. J. Geophys. Res. Atmos. 2016, 121, 5699–5717. [Google Scholar] [CrossRef]
- Mao, K.; Shi, J.; Li, Z.; Qin, Z.; Li, M.; Xu, B. A Physics-based statistical algorithm for retrieving land surface temperature from amsr-e passive microwave data. Sci. China Ser. Earth Sci. 2007, 50, 1115–1120. [Google Scholar] [CrossRef]
- Parinussa, R.; Lakshmi, V.; Johnson, F.; Sharma, A. Comparing and combining remotely sensed land surface temperature products for improved hydrological applications. Remote Sens. 2016, 8, 162. [Google Scholar] [CrossRef] [Green Version]
Station ID | Longitude (°) | Latitude (°) | Land Cover Type | Climate Type | Sensor Viewing Height (m) | Data Cover Time |
---|---|---|---|---|---|---|
Qianyanzhou | 115.058 | 26.741 | Evergreen needleleaf forest | Subtropical monsoon climate | 41 | 2013–2016 |
Dinghushan | 112.534 | 23.173 | Evergreen broadleaf forest | Humid monsoon climate | 27 | 2014–2015 |
Taoyuan | 111.411 | 28.897 | Croplands | Subtropical humid monsoon climate | 2 | 2014 |
Xishuangbanna | 101.265 | 21.928 | Evergreen broadleaf forest | Tropical monsoon climate | 48.8 | 2014–2015 |
Flux Sites | Condition | Daytime | Nighttime | ||||
---|---|---|---|---|---|---|---|
MAE (K) | RMSE (K) | R2 | MAE (K) | RMSE (K) | R2 | ||
Qianyanzhou | MODIS | 1.86 | 2.36 | 0.96 | 2.47 | 2.81 | 0.92 |
Clear sky | 1.79 | 2.45 | 0.96 | 2.69 | 3.01 | 0.92 | |
Cloud sky | 1.98 | 2.76 | 0.94 | 2.62 | 3.05 | 0.89 | |
Dinghushan | MODIS | 2.49 | 3.10 | 0.85 | 2.41 | 3.03 | 0.85 |
Clear sky | 2.53 | 3.24 | 0.83 | 3.11 | 3.55 | 0.86 | |
Cloud sky | 2.69 | 3.32 | 0.81 | 2.60 | 3.17 | 0.82 | |
Taoyuan | MODIS | 2.73 | 3.38 | 0.89 | 1.66 | 2.12 | 0.92 |
Clear sky | 3.58 | 3.63 | 0.88 | 1.78 | 2.38 | 0.92 | |
Cloud sky | 2.81 | 3.29 | 0.90 | 2.05 | 2.62 | 0.91 | |
Xishuangbanna | MODIS | 3.17 | 3.66 | 0.84 | 2.51 | 2.98 | 0.87 |
Clear sky | 2.76 | 3.33 | 0.82 | 2.21 | 2.80 | 0.84 | |
Cloud sky | 2.48 | 3.11 | 0.79 | 2.25 | 2.77 | 0.86 |
Stations | Land Cover Types | Thermal Heterogeneity (K) | Clear Skies | Cloudy Skies | ||||
---|---|---|---|---|---|---|---|---|
MAE (K) | RMSE (K) | R2 | MAE (K) | RMSE (K) | R2 | |||
1 | Croplands | 0.91 | 2.10 | 2.65 | 0.95 | 2.41 | 3.14 | 0.88 |
2 | Croplands/natural vegetation mosaic | 0.66 | 1.96 | 2.52 | 0.94 | 2.13 | 2.90 | 0.88 |
3 | Savannas | 0.84 | 1.73 | 2.23 | 0.96 | 2.20 | 2.91 | 0.90 |
4 | Croplands | 0.99 | 2.28 | 2.94 | 0.93 | 2.56 | 3.30 | 0.90 |
5 | Urban and built-up | 0.85 | 2.11 | 2.71 | 0.95 | 2.67 | 3.80 | 0.84 |
6 | Woody savannas | 1.00 | 2.14 | 2.71 | 0.92 | 2.26 | 2.97 | 0.87 |
7 | Evergreen broadleaf forest | 0.62 | 2.12 | 2.77 | 0.86 | 2.57 | 3.40 | 0.77 |
8 | Urban and built-up | 0.97 | 2.14 | 2.75 | 0.95 | 2.58 | 3.39 | 0.89 |
9 | Grasslands | 0.93 | 1.71 | 2.22 | 0.93 | 1.99 | 2.58 | 0.85 |
10 | Savannas | 0.71 | 1.94 | 2.49 | 0.94 | 2.00 | 2.64 | 0.88 |
Stations | Products | Bias (K) | MAE (K) | RMSE (K) | R2 |
---|---|---|---|---|---|
Qianyanzhou | MODIS (10 km) | −0.52 | 1.64 | 2.11 | 0.96 |
MLP (10 km) | −0.93 | 1.85 | 2.45 | 0.94 | |
Dinghushan | MODIS (10 km) | 0.40 | 1.89 | 2.29 | 0.93 |
MLP (10 km) | −0.58 | 1.61 | 2.14 | 0.92 | |
Taoyuan | MODIS (10 km) | −3.09 | 3.49 | 4.00 | 0.94 |
MLP (10 km) | −3.28 | 3.80 | 4.54 | 0.90 | |
Xishuangbanna | MODIS (10 km) | 0.76 | 1.86 | 2.39 | 0.88 |
MLP (10 km) | 0.25 | 1.90 | 2.44 | 0.84 |
Metrics | Qianyanzhou | Dinghushan | Taoyuan | Xishuangbanna | ||||
---|---|---|---|---|---|---|---|---|
GTWR | One-Step | GTWR | One-Step | GTWR | One-Step | GTWR | One-Step | |
Bias (K) | 0.06 | −0.37 | 0.12 | −0.60 | 0.27 | −0.94 | −1.90 | −2.23 |
MAE (K) | 2.17 | 2.59 | 2.35 | 2.27 | 2.12 | 2.41 | 2.35 | 2.95 |
RMSE (K) | 2.73 | 3.23 | 2.90 | 2.93 | 2.82 | 3.27 | 2.91 | 3.56 |
R2 | 0.91 | 0.88 | 0.85 | 0.86 | 0.95 | 0.91 | 0.87 | 0.86 |
Publisher’s Note: MDPI stays neutral with regard to jurisdictional claims in published maps and institutional affiliations. |
© 2021 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (http://creativecommons.org/licenses/by/4.0/).
Share and Cite
Gao, Z.; Hou, Y.; Zaitchik, B.F.; Chen, Y.; Chen, W. A Two-Step Integrated MLP-GTWR Method to Estimate 1 km Land Surface Temperature with Complete Spatial Coverage in Humid, Cloudy Regions. Remote Sens. 2021, 13, 971. https://doi.org/10.3390/rs13050971
Gao Z, Hou Y, Zaitchik BF, Chen Y, Chen W. A Two-Step Integrated MLP-GTWR Method to Estimate 1 km Land Surface Temperature with Complete Spatial Coverage in Humid, Cloudy Regions. Remote Sensing. 2021; 13(5):971. https://doi.org/10.3390/rs13050971
Chicago/Turabian StyleGao, Zhen, Ying Hou, Benjamin F. Zaitchik, Yongzhe Chen, and Weiping Chen. 2021. "A Two-Step Integrated MLP-GTWR Method to Estimate 1 km Land Surface Temperature with Complete Spatial Coverage in Humid, Cloudy Regions" Remote Sensing 13, no. 5: 971. https://doi.org/10.3390/rs13050971