Energy-conservation datasets of global land surface radiation and heat fluxes from 2000-2020 generated by CoSEB

Land surface radiation balance and heat balance play an important role in Earth's climate system, representing the physical processes by which the surface-atmosphere absorb and redistribute the radiation and heat fluxes, and facilitating the exchange of water, energy, carbon.The datasets provide globally gridded, energy-conserving estimates of eight land surface radiation and heat fluxes, including downward shortwave radiation (SW_IN), downward longwave radiation (LW_IN), upward shortwave radiation (SW_OUT), upward longwave radiation (LW_OUT), net radiation (Rn), latent heat flux (LE), soil heat flux (G) and sensible heat flux (H). The data span February 26, 2000 to December 31, 2020, at a daily temporal resolution and 0.05° spatial resolution.The datasets were generated by Coordinated estimates of global land Surface Energy Balance components (CoSEB) model that was renewed with a combination of in-situ observations from FLUXNET, AmefiFlux, EuroFlux, OzFlux, TPDC and ChinaFLUX, remote sensing data from GLASS and MODIS, and ERA5-Land reanalysis datasets.The eight global land surface radiation and heat fluxes included in the datasets comply with both energy balance and radiation balance. the root mean square errors (RMSEs) for daily estimates of SW_IN, SW_OUT, LW_IN, LW_OUT, Rn, LE, H and G from the CoSEB-based datasets are 28.51 W/m², 10.39 W/m², 14.29 W/m², 10.62 W/m², 22.40 W/m², 24.38 W/m², 22.67 W/m² and 6.77 W/m², respectively, when validated against in-situ observations at 134 EC sites.The data is stored in GeoTIFF format with a file name of <VARIABLE>-YYYYMMDD.tif, where <VARIABLE> represents the ID of the variables, YYYY represents the year, MM represents the month and DD represents the day, e.g. LW_IN-20000226.tif represents the value of downward longwave radiation on February 26th, 2000. The unit of each variable is MJ/(m² day). Note that the invalid value is -9999, and the value of each variable = the value of each pixel * scale factor, where scale factor is 0.01.