Skip to Content
Remote SensingRemote Sensing
  • Article
  • Open Access

7 October 2017

33 Pages

A Modified Multi-Source Parallel Model for Estimating Urban Surface Evapotranspiration Based on ASTER Thermal Infrared Data

,
,
,
,
,
and
1
School of Environmental Science and Spatial Informatics, China University of Mining and Technology, Xuzhou 221116, China
2
Department of Geography, Earth System Science, Vrije Universiteit Brussel, Brussels 1050, Belgium
3
Key Laboratory of Digital Earth Science, Institute of Remote Sensing and Digital Earth, Chinese Academy of Sciences, Beijing 100094, China
4
School of Geography, Geomatics and Planning, Jiangsu Normal University, Xuzhou 221116, China

Abstract

To date, little attention has been given to remote sensing-based algorithms for inferring urban surface evapotranspiration. A multi-source parallel model based on ASTER data was one of the first examples, but its accuracy can be improved. We therefore present a modified multi-source parallel model in this study, which has made improvements in parameterization and model accuracy. The new features of our modified model are: (1) a characterization of spectrally heterogeneous urban impervious surfaces using two endmembers (high- and low-albedo urban impervious surface), instead of a single endmember, in linear spectral mixture analysis; (2) inclusion of an algorithm for deriving roughness length for each land surface component in order to better approximate to the actual land surface characteristic; and (3) a novel algorithm for calculating the component net radiant flux with a full consideration of the fraction and the characteristics of each land surface component. HJ-1 and ASTER data from the Chinese city of Hefei were used to test our model’s result with the China–ASEAN ET product. The sensitivity of the model to vegetation and soil fractions was analyzed and the applicability of the model was tested in another built-up area in the central Chinese city of Wuhan. We conclude that our modified model outperforms the initial multi-source parallel model in accuracy. It can obtain the highest accuracy when applied to vegetation-dominated (vegetation proportion > 50%) areas. Sensitivity analysis shows that vegetation and soil fractions are two important parameters that can affect the ET estimation. Our model is applicable to estimate evapotranspiration in other urban areas.

Article Metrics

Citations

Article Access Statistics

Multiple requests from the same IP address are counted as one view.