Performance Metrics for Soil Moisture Downscaling Methods: Application to DISPATCH Data in Central Morocco

Abstract

Data disaggregation (or downscaling) is becoming a recognized modeling framework to improve the spatial resolution of available surface soil moisture satellite products. However, depending on the quality of the scale change modeling and on the uncertainty in its input data, disaggregation may improve or degrade soil moisture information at high resolution. Hence, defining a relevant metric for evaluating such methodologies is crucial before disaggregated data can be eventually used in fine-scale studies. In this paper, a new metric, named G DOWN , is proposed to assess the potential gain provided by disaggregation relative to the non-disaggregation case. The performance metric is tested during a four-year period by comparing 1-km resolution disaggregation based on physical and theoretical scale change (DISPATCH) data with the soil moisture measurements collected by six stations in central Morocco. DISPATCH data are obtained every 2-3 days from 40-km resolution SMOS (Soil Moisture Ocean Salinity) and 1-km resolution optical MODIS (Moderate Resolution Imaging Spectroradiometer) data. The correlation coefficient between G DOWN and the disaggregation gain in time series correlation, mean bias and bias in the slope of the linear fit ranges from 0.5 to 0.8. The new metric is found to be a good Remote Sens. 2015, 7 3784 indicator of the overall performance of DISPATCH. Especially, the sign of G DOWN (positive in the case of effective disaggregation and negative in the opposite case) is independent of the uncertainties in SMOS data and of the representativeness of localized in situ measurements at the downscaling (1 km) resolution. In contrast, the traditional root mean square difference between disaggregation output and in situ measurements is poorly correlated (correlation coefficient of about 0.0) with the disaggregation gain in terms of both time series correlation and bias in the slope of the linear fit. The G DOWN approach is generic and thus could help test a range of downscaling methods dedicated to soil moisture and to other geophysical variables.

Stochastic Bias Correction and Uncertainty Estimation of Satellite-Retrieved Soil Moisture Products

“…Additionally, the international networks still contain upscaling or representativeness errors, and have a limited global coverage, as they are just a few pixels out of the whole at a hundred thousand pixels [22]. On the other hand, downscaling that disaggregates the satellite products to a local scale is also complicated [32,33]. The currently available downscaling techniques often employ machine learning or proxies such as land surface temperature, or vegetation index for downscaling soil moisture products at low resolution.…”

“…On the other hand, downscaling that disaggregates the satellite products to a local scale is also complicated [32,33]. The currently available downscaling techniques often employ machine learning or proxies such as land surface temperature, or vegetation index for downscaling soil moisture products at low resolution.…”

“…For example, an instantaneously or seasonally dynamic variation of rescaling parameters or retrieval errors is ignored [53]. This method is based on soil moisture climatology, which do not take into account sub-pixel heterogeneity [33]. Thus, satellite observational or retrieval errors often remain even after bias correction [54].…”

“…Vegetation, meteorological activity, topography, and soil property are spatially heterogeneous at low-resolution [33,40,83], although satellite retrievals deal with the pixels as a uniform entity (the SMOS algorithm considers sub-pixel land cover heterogeneity at 4 km by 4 km, but it is eventually aggregated to be uniform at 25 km by 25 km). Crow et al [59] suggested that aggregation errors arise from the nonlinearity between soil moisture retrievals and land surface, and from the space borne sensors that are unable to capture the net impact of sub-pixel land surface conditions.…”

Temporal and Spatial Comparison of Agricultural Drought Indices from Moderate Resolution Satellite Soil Moisture Data over Northwest Spain

“…During the last several years, there has been a growing interest in developing techniques to enhance the spatial resolution of SMOS and SMAP SM observations. For SMOS, pixel disaggregation approaches are usually based on the synergy of passive microwave and optical visible/infrared (VIS/IR) observations [57][58][59][60][61][62]. Currently, the SMOS Barcelona Expert Centre (BEC) Level 4 (L4) SM maps at 1 km spatial resolution over the Iberian Peninsula are freely distributed (http://bec.icm.csic.es).…”