Mapping surface heat fluxes by assimilating GOES land surface temperature and SMAP products

Surface heat fluxes significantly affect the land-atmosphere interaction, but their modelling is often hindered by the lack of in-situ measurements and the high spatial heterogeneity. Here, we propose a hybrid particle assimilation strategy to estimate surface heat fluxes by assimilating GOES land surface temperature (LST) data and SMAP products into a simple dual-source surface energy balance model, in which the requirement for in-situ data is minimized. The study aims to estimate two key parameters: a neutral bulk heat transfer coefficient (CHN) and an evaporative fraction (EF). CHN scales the sum of surface energy fluxes, and EF represents the partitioning between flux components. To bridge the huge resolution gap between GOES and SMAP data, SMAP data are assimilated using a particle filter to update soil moisture which constrains EF, and GOES data are assimilated with an adaptive particle batch smoother to update CHN. The methodology is applied to an area in the US Southern Great Plains with forcing data from NLDAS-2 and the GPM mission. Assessment against in-situ observations suggests that the sensible and latent heat flux estimates are greatly improved at both daytime and 30-min scale after assimilation, particularly for latent heat fluxes. Comparison against an LST-only assimilation case demonstrates that despite the coarse resolution, assimilating SMAP data is not only beneficial but also crucial for successful and robust flux estimation, particularly when the modelling uncertainties are large. Since the methodology is independent on in-situ data, it can be easily applied to other areas.