A Method for Upscaling In Situ Soil Moisture Measurements to Satellite Footprint Scale Using Random Forests

Abstract

Geophysical products generated from remotely sensed data require validation to evaluate their accuracy. Typically in situ measurements are used for validation, as is the case for satellite-derived soil moisture products. However, a large disparity in scales often exists between in situ measurements (covering meters to 10s of meters) and satellite footprints (often hundreds of meters to several kilometers), making direct comparison difficult. Before using in situ measurements for validation, they must be 'upscaled' to provide the mean soil moisture within the satellite footprint. There are a number of existing upscaling methods previously applied to soil moisture measurements, but many place strict requirements on the number and spatial distribution of soil moisture sensors difficult to achieve with permanent/semi-permanent ground networks necessary for long-term validation efforts. A new method for upscaling is presented here, using Random Forests to fit a model between in situ measurements and a number of landscape parameters and variables impacting the spatial and temporal distributions of soil moisture. The method is specifically intended for validation of the NASA Soil Moisture Active Passive (SMAP) products at 36, 9 and 3 km scales. The method was applied to in situ data from the SoilSCAPE network in California, validated with data from the SMAPVEX12 campaign in Manitoba, Canada with additional verification from the TxSON network in Texas. For the SMAPVEX12 site the proposed method was compared to extensive field measurements and was able to predict mean soil moisture over a large area more accurately than other upscaling approaches.

The Texas Soil Observation Network:A Comprehensive Soil Moisture Dataset for Remote Sensing and Land Surface Model Validation

“…The objective of TxSON was to establish a spatially representative measure of SWC for the calibration and validation of remotely sensed estimates (Chan et al, 2016, 2018; Colliander et al, 2017b, 2018; Kim et al, 2017; Ouellette et al, 2017; Bindlish et al, 2018; Das et al, 2018), upscaling exercises (Clewley et al, 2017), and data assimilation and land surface model validation (Kolassa et al, 2017, 2018; Reichle et al, 2017). Much of the past work has focused on SMAP products that are posted to the Equal‐Area Scalable Earth Version 2 (EASE2) grid.…”

“…The arithmetic average (Method 1) was calculated for unflagged data from each in situ location within the given EASE2 pixel (Table 2) at 36, 9, and 3 km at each depth and the SD calculated for each hour. This approach was evaluated by others and found to be robust, particularly when replication was sufficient within the spatial extent (Adams et al, 2015; Clewley et al, 2017; Bhuiyan et al, 2018).…”

Synergetic Use of Sentinel-1 and Sentinel-2 Data for Soil Moisture Mapping at Plot Scale

“…Lower-resolution soil moisture information is required for applications related to natural disasters, such as flooding or land-slides (early warning), but also in the agricultural sector to increase irrigation and fertilization efficiency. Due to this fact, a considerable interest has been shown in downscaling and upscaling methods in recent years [35,[39][40][41][42]. In this study, the synergy of single-polarized C-band Sentinel-1 and optical Sentinel-2 data was used to retrieve soil moisture at the plot scale in vegetated areas.…”

Recovery of Soil Moisture Active Passive (SMAP) Instrument's Active Measurements via Coupled Dictionary Learning

“…where µ and σ stand for the mean value and the standard deviation, respectively. Finally, we measured the absolute difference [9] between the reconstructed and the ground truth active measurements. The absolute difference is widely used in soil moisture retrieval problems.…”

“…The authors use physics-based models in order to couple the scattering and emission processes.Additionally, they define a joint cost function that balances the contributions of radar's and radiometer's measurements, extracting the optimal estimates over a larger range of surface soil moisture. Finally, the authors in [9] present an efficient machine learning scheme, relied on Random Forests that upscales in-situ soil moisture estimates to satellite footprints scale of SMAP, in order to validate with high accuracy the soil moisture retrieval.…”