Calibration and Validation
Problem statement: Land Surface Temperature (LST) is one of the main quantities governing the energy exchange between surface and atmosphere. On the extensive Tibetan Plateau (TP), where in-situ observations are usually extremely sparse, accurate knowledge of the land surface energy balance is crucial for understanding and simulating regional processes of meteorology, hydrology and ecology. More specifically, all-weather LST products are required for accurately simulating soil heat transfer, which provides insights into changes in TP permafrost / seasonally frozen ground and regional climate change. However, LST products based on thermal infrared (TIR) remote sensing are limited to clear sky conditions. Recently two all-weather satellite LST products became available, but they still require more extensive validation and assessment of their uncertainty. Objectives: The main objective is to inter-compare and validate the two new LST products, which provide (nearly) gap-free all-weather LST at high spatial resolution. The two all-weather LST products utilise different retrieval approaches, namely the method by – Zhang et al. (2019): temporal component decomposition and merging of TIR LST with passive microwave (PMW) LST. – Martins et al. (2019): merging of clear-sky MSG/SEVIRI LST with LST generated by a Soil-Vegetation-Atmosphere (SVAT) model under cloudy conditions. Further objectives: – Generation of long term (global) all-weather LST data set – Setting up an LST validation station in China to provide Fiducial Reference Measurements (FRM) – Employing all-weather LST data to simulate and study freeze / thaw on the TP Method: The two new all-weather LST products and LST extracted from ERA5-Land data, which are provided by Copernicus Climate Change Service (C3S), will be inter-compared over selected regions in China, Europe, and Southern Africa. All inter-comparisons will utilise ESA GlobTemperature (GT) harmonised data format (netCDF) and their relative performance will be assessed to provide insights into their respective strengths and limitations. The three LST products will be validated against in-situ measurements from the following station networks: 1) Karlsruhe Institute of Technology (KIT), 2) Baseline Surface Radiation Network (BSRN), 3) European Fluxes Database Cluster (EFDC) initiative, 4) Heihe Watershed Allied Telemetry Experimental Research (HiWATER) and Watershed Allied Telemetry Experimental Research (WATER) in the Heihe River basin, and 5) networks operated by other Chinese groups on the TP. Based on experience and an instrument package provided by KIT, the Chinese partners will set up a new LST validation station in China. Since the thermal sampling depth correction (TSDC) between TIR LST and PMW LST is larger for dry soil, all-weather LST determined over the arid validation site Gobabeb (Namibia) will be compared with results from the Zhou et al. (2017) method, which explicitly models soil heat conduction. The main causes of differences between the three LST products will be identified and used to improve the estimates of LSA SAF all-weather LST uncertainty. The all-weather LST retrieved from merged TIR and PWM data will serve as input for model simulations of freeze / thaw on the TP. Deliverables: – Inter-comparison and validation results for the two all-weather LST products – Assessment of all-weather LST product uncertainties – Results from simulating freeze / thaw on TP Source of funding: Frank-M. Göttsche is funded by Karlsruhe Institute of Technology (KIT). João P.A. Martins is funded by Instituto Português do Mar e da Atmosfera /Portuguese Institute for the Sea and the Atmosphere (IPMA). Ji Zhou is funded by the National Natural Science Foundation of China under Grant 41871241 and the University of Electronic Science and Technology of China (UESTC). Wenjiang Zhang is funded by the National Natural Science Foundation of China under Grant 41771112.
