By Topic

Studies on the Relationships Between Land Surface Temperature and Environmental Factors in an Inland River Catchment Based on Geographically Weighted Regression and MODIS Data

Sign In

Cookies must be enabled to login.After enabling cookies , please use refresh or reload or ctrl+f5 on the browser for the login options.

Formats Non-Member Member
$33 $13
Learn how you can qualify for the best price for this item!
Become an IEEE Member or Subscribe to
IEEE Xplore for exclusive pricing!
close button

puzzle piece

IEEE membership options for an individual and IEEE Xplore subscriptions for an organization offer the most affordable access to essential journal articles, conference papers, standards, eBooks, and eLearning courses.

Learn more about:

IEEE membership

IEEE Xplore subscriptions

5 Author(s)
Fei Tian ; College of Resources Science and Technology, Beijing Normal University, Beijing, China ; Guo Yu Qiu ; Yong Hui Yang ; Yu Jiu Xiong
more authors

Despite growing concerns in Land Surface Temperature (LST) and its related environmental factors (geographical, climate, and atmospheric conditions), little attention was about the spatial variation that consider above conditions together. Our purpose is to analyze and quantify LST and related environmental factors, using Geographically Weighted Regression (GWR), and Moderate Resolution Imaging Spectroradiometer (MODIS) data in a typical inland river catchment, named Heihe River catchment, China. Considering thirteen environmental factors (altitude, latitude, Topographic Wetness Index, Cos(aspect), temperature, precipitation, humidity, wind speed, radiation, albedo, the normalized difference vegetation index (NDVI), water vapor, COT), 18 GWR models were set up. Results showed that yearly averaged LST changed from 264 K to 309 K, with the highest value recorded in the downstream desert region. LST has the same variable trend and seasonality with NDVI, precipitable water vapor, and cloud optical thickness (COT), but has an inverse relationship with albedo. All GWR models indicated better simulation with smaller Akaike Information Criterion (AICc), and higher coefficient of determination (R2), compared with Ordinary Least Squares method (OLS). Furthermore, performance of multi-factor analysis was better than single-factor analysis, with model 18 showing the best performance achieving higher R2 (0.94) and lower AICc (7760). For all GWR model, 86.4% of R2 was higher than 0.60, most values distributed in the range of 0.80-0.99, and 86.59% of residual values were within the range of ±2 K. Different parameters resulted in different slope distribution, which indicated that altitude is the major driving factor, followed by NDVI, and albedo.

Published in:

IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing  (Volume:5 ,  Issue: 3 )