. L-band High Spatial Resolution Soil Moisture Mapping using Small Unmanned Aerial Systems

Abstract
Soil moisture is of fundamental importance to many hydrological, biological, and geochemical processes. It plays an important role in the development and evolution of convective weather and precipitation and impacts water resource management, agriculture, and flood runoff prediction. The launch of NASA's Soil Moisture Active Passive (SMAP) mission in 2015 has provided global measurements of soil moisture and surface freeze/thaw state at fixed crossing times and spatial resolutions as small as 9 km for some products. However, there exists a need for measurements of soil moisture on much smaller spatial scales and arbitrary diurnal times for satellite data validation, precision agriculture, and evaporation and transpiration studies of boundary layer heat transport. The Lobe Differencing Correlation Radiometer (LDCR) developed by the University of Colorado's Center for Environmental Technology (CET) on Black Swift Technologies (BST) LLC small unmanned aerial system (sUAS) provides a new mean of airborne mapping of soil moisture on spatial scales as small as several meters (i.e., the height of the aircraft platform). The LDCR performance was validated during flight tests at the Canton, Oklahoma Soilscape site in September 2015, and again at the Irrigation Research Foundation (IRF) in Yuma Colorado in June 2016. The upwelling antenna temperature TA and surface infrared physical temperature TP were measured during several ~hour-long sorties. These data provided primary inputs to a soil moisture mapping algorithm based on a soil vegetation radiative transfer (SVRT) model. However, there are several challenging aspects to the use of sUAS L-band data: (i) sUAS flight lines are irregular due to wind variations, line of site flight requirement, obstacle avoidance, and other operational considerations, (ii) the desired regular mapping grid is typically more dense than the flight pattern, and (iii) the vegetation cover is often inhomogeneous at the scale of the available observations. To accommodate these unique spatial attributes, a new high spatial resolution soil moisture mapping algorithm was developed specifically for sparse irregular sUAS observations. The LDCR TA data at sUAS sampling grid is subsequently mapped to volumetric soil moisture (VSM) on a user-defined grid using a full-domain linear minimum mean square error (LMMSE) estimation method. Various spatial covariance functions of VSM are addressed in this retrieval as relevant for managed cropland of various vegetation types. Initially retrieved LDCR VSM data are compared favorably with in-situ measured VSM data and irrigation records.