Sub-metre mapping of surface soil moisture in proglacial valleys of the tropical Andes using a multispectral unmanned aerial vehicle

Abstract

Surface soil moisture is a critical but often neglected component of the hydrologic budget. Within mountain environments, surface soil moisture is highly heterogeneous and challenging to measure. Point measurements are often poorly representative of larger areas, while satellite pixels are generally too coarse in these topographically varied landscapes. In the Cordillera Blanca, Peru, rapid glacier recession is impacting downstream water supply in timing, quantity and quality. Recent research has shown that these proglacial valleys provide vital ecosystem services and store considerable amounts of water within the groundwater systems. However, the spatial and temporal variability of soil water storage is poorly understood. In this tropical Andean setting, we use an unmanned aerial vehicle (UAV) with multispectral (visible, near infrared, thermal infrared) sensors to map sections of two proglacial valleys in the Cordillera Blanca, Peru at sub-metre resolution. We use the Temperature Vegetation Dryness Index (TVDI) (Sandholt et al., 2002), and apply it for the first time to UAV borne imagery to estimate absolute surface soil moisture through calibration with in-scene field measurements. Resulting surface soil moisture maps have 50 cm spatial resolution with an R2 value of 0.55 and 0.76 for the two study sites, Pachacoto and Llanganuco respectively. We analyse the multispectral orthomosaics and soil moisture maps to improve our understanding of the controls on spatial variability in surface soil moisture within the proglacial valleys of the Cordillera Blanca. The maps permit us to identify both important groundwater spring sources and areas where domestic grazing impacts observed soil moisture estimates. The high resolution observations provided by the UAV facilitate a greater understanding of fine scale heterogeneity within these environments.