Estimating evaporative fraction from readily obtainable variables in mangrove forests of the Everglades, U.S.A.

International Journal of Remote Sensing
By: , and 



A remote-sensing-based model to estimate evaporative fraction (EF) – the ratio of latent heat (LE; energy equivalent of evapotranspiration –ET–) to total available energy – from easily obtainable remotely-sensed and meteorological parameters is presented. This research specifically addresses the shortcomings of existing ET retrieval methods such as calibration requirements of extensive accurate in situ micrometeorological and flux tower observations or of a large set of coarse-resolution or model-derived input datasets. The trapezoid model is capable of generating spatially varying EF maps from standard products such as land surface temperature (Ts) normalized difference vegetation index (NDVI) and daily maximum air temperature (Ta). The 2009 model results were validated at an eddy-covariance tower (Fluxnet ID: US-Skr) in the Everglades using Ts and NDVI products from Landsat as well as the Moderate Resolution Imaging Spectroradiometer (MODIS) sensors. Results indicate that the model accuracy is within the range of instrument uncertainty, and is dependent on the spatial resolution and selection of end-members (i.e. wet/dry edge). The most accurate results were achieved with the Ts from Landsat relative to the Tfrom the MODIS flown on the Terra and Aqua platforms due to the fine spatial resolution of Landsat (30 m). The bias, mean absolute percentage error and root mean square percentage error were as low as 2.9% (3.0%), 9.8% (13.3%), and 12.1% (16.1%) for Landsat-based (MODIS-based) EF estimates, respectively. Overall, this methodology shows promise for bridging the gap between temporally limited ET estimates at Landsat scales and more complex and difficult to constrain global ET remote-sensing models.

Publication type Article
Publication Subtype Journal Article
Title Estimating evaporative fraction from readily obtainable variables in mangrove forests of the Everglades, U.S.A.
Series title International Journal of Remote Sensing
DOI 10.1080/01431161.2017.1312033
Volume 38
Issue 14
Year Published 2017
Language English
Publisher Taylor & Francis
Contributing office(s) Eastern Geographic Science Center
Description 27 p.
First page 3981
Last page 4007
Google Analytics Metrics Metrics page
Additional publication details