Deriving Daily Time Series Evapotranspiration, Evaporation and Transpiration Maps With Landsat Data

Abstract:

Mapping high resolution evapotranspiration (ET) over large region at daily time step is complex and computationally intensive. Utility of high resolution daily ET maps are large ranging from crop water management to watershed management. The aim of this work is to generate daily time series (10 years) ET and its components vegetation transpiration (T) and soil water evaporation (E) maps using Landsat 5 satellite data for Southern Great Plains forage–rangeland–winter wheat production system in Oklahoma (OK). Framework for generating these products included the two source energy balance (TSEB) algorithm and other important features were: (a) atmospheric correction algorithm; (b) spatially interpolated weather inputs; (c) functions for varying Priestley-Taylor coefficient; and (d) ET, E and T extrapolating algorithm utilizing reference ET. An extensive network of 140 weather stations managed by Oklahoma Mesonet was utilized to generate spatially interpolated inputs of air temperature, relative humidity, wind speed, solar radiation, pressure, and reference ET. Validation of the ET maps were done against eddy covariance data from two grassland sites at El Reno, OK suggested good performance (Table 1). Figure 1 illustrates a daily ET map for a very small subset of 18^thJuly 2006 ET map, where difference in ET among different land uses such as the irrigated cropland, vegetation along drainage, and grassland is very distinct. Results indicated that the proposed ET mapping framework is suitable for deriving high resolution time series daily ET maps at regional scale with Landsat Thematic Mapper data.

.

Table 1: Daily actual ET performance statistics for two grassland locations at El Reno OK for year 2005

.

Management Type	Mean (obs) (mm d-1)	Mean (est)  (mm d-1)	MBE (mm d-1)	% MBE (%)	RMSE (mm d-1)	RMSE (%)	MAE (mm d-1)	MAPD (%)	NSE	R2
Control	2.2	1.8	-0.43	-19.4	0.87	38.9	0.65	29.5	0.71	0.79
Burnt	2.0	1.8	-0.15	-7.7	0.80	39.8	0.62	30.7	0.73	0.77