Stream Cold Water Patches identification and characterisation: acquisition and analysis approaches using Uncrewed Aerial Systems (UASs)-based imagery

Abstract

Stream cold water patches (CWPs) play an increasingly crucial role in the persistence of cold-adapted biological communities and the resilience of river ecosystems, particularly in the context of global warming. Predicting the spatial distribution of CWPs is key to understanding their drivers and informing river management strategies targeting ecosystem resilience. Uncrewed Aerial Systems (UAS)-based Thermal Infra-Red (TIR) imagery technologies provide a novel tool to identify and thermally quantify CWPs at large spatial scales. When used in conjunction with optical (RGB) imagery, CWPs physical attributes and their surrounding riverscape can be characterised, allowing to assess drivers of potential thermal refugia at multiple spatial scales. However, several critical challenges in UAS-based data acquisition, processing and interpretation must be addressed. They include a requirement of perfect TIR-RGB overlapping during data acquisition to enable the identification and extraction of true water temperature pixels, in-situ stream temperature measurements to enable calibration and validation during post-processing, and the definitions of thermal and areal metric thresholds based on targeted ecological processes to identify, characterise and interpret CWPs distributions. We developed a set of methodological approaches used to address these challenges and obtain reliable CWPs information from UAS-based data as a basis to promote effective management of present and future freshwater ecosystems.