Validation of remotely sensed evapotranspiration over the Hai River Basin, China

Abstract

[1] Ground-based validation is crucial for ensuring the accuracy of remotely sensed evapotranspiration (RS_ET) and extending its application. This paper proposes an innovative validation method based on multisource evapotranspiration (ET) from ground measurements, with the validation results including the accuracy assessment, error source analysis, and uncertainty analysis of the validation process. It is a potentially useful approach to evaluate the accuracy and analyze the spatiotemporal properties of RS_ET at both the basin and local scales, and is appropriate to validate RS_ET in diverse resolutions at different time-scales. An independent RS_ET validation using such a method was presented over the Hai River Basin in 2002-2009, China. In general, validation at the basin scale showed good agreements between the 1 km annual RS_ET and the validation data such as the water balance ET (root-mean-square error (RMSE): 50.73 mm), MODIS ET products (RMSE: 79.84 mm), precipitation, and land use types. At the local scale, multiscale ET measurements from large aperture scintillometer (LAS) and eddy covariance system (EC) with a footprint model were used for validation over three typical landscapes. In most cases, the 1 km RS_ET resulted in slight overestimation with the LAS measurements (RMSE: 10.75 mm for monthly results, 0.78 mm for daily results), while the 30 m RS_ET was underestimated compared to the EC measurements (RMSE: 16.28 mm for monthly results, 0.99 mm for daily results). Furthermore, error sources of RS_ET and uncertainties of the validation process were investigated in detail. The results showed that the proposed validation method was reasonable and feasible.

Validation of Global Evapotranspiration Product (MOD16) using Flux Tower Data in the African Savanna, South Africa

“…In addition, the layout of a single 1 km pixel may not directly match the flux tower footprints and may add further spatial discrepancy between the two. Footprint modelling is a means to reduce the spatial discrepancy between flux tower measurements and MODIS pixel [36], but this was beyond the scope of this study.…”

“…Kim et al [44] later validated the global product using Asiaflux stations. Jia et al [36] evaluated spatiotemporal MODIS ET product in the Hai river basin. Inaccuracies, such as over-or underestimation, and no relationship were observed between the flux tower and MOD16 ET estimates in the above publications.…”

“…• empirical methods that involve the use of statistically-derived relationships between ET and vegetation indices such as the normalized difference vegetation index (NDVI) or the enhanced vegetation index (EVI) [20][21][22][23][24][25], • residual methods of surface energy balance (single-and dual-source models) [8,26] which include the Surface Energy Balance Algorithm over Land (SEBAL) [27,28], Surface Energy Balance System (SEBS) [8,29,30] and Mapping EvapoTranspiration at high Resolution with Internalized Calibration (METRIC) [6,31,32], • physically-based methods that involve the application of the combination of Penman-Monteith [7,33,34] and Priestley-Taylor types of equations [35][36][37][38][39], and • Data assimilation methods adjoined to the heat diffusion equation [40] and through the radiometric surface temperature sequences [41].…”

Evaluation and Aggregation Properties of Thermal Infra-Red-Based Evapotranspiration Algorithms from 100 m to the km Scale over a Semi-Arid Irrigated Agricultural Area

“…The four aggregation rules of roughness length (linear, geometric i.e., Equation (25), harmonic i.e., Equation (26) and conductance average i.e., Equation (30)) were implemented to derive input roughness lengths for SEBS at LR. The resulting ET calculated by SEBS was compared to the 'agg-ASTER' ET fluxes.…”

Comparative Assessment of Two Vegetation Fractional Cover Estimating Methods and Their Impacts on Modeling Urban Latent Heat Flux Using Landsat Imagery

“…These variables were computed as follows: (8) where e a is the atmospheric water vapor pressure in hPa, e o is the saturation vapor pressure in hPa and is calculated using Equations (9) and (10) [5,7]. γ is the psychometric constant.…”