This dataset is produced on a 1°x1° global grid using satellite-derived cloud parameters, ozone fields from multiple sources, reanalysis meteorology, and several other ancillary datasets. Both SW and LW surface fluxes are derived with two sets of algorithms: one designated as primary, and the other as quality-check. TOA fluxes are derived with primary algorithms only. All except the quality-check SW algorithm derive fluxes on a 3-hourly time resolution, which are then averaged into daily, monthly, and monthly/3-hourly values. Quality-check SW fluxes are derived on a daily resolution and averaged into monthly values. This dataset is designated as SRB Release-3.0 and supersedes all earlier releases. Surface fluxes from all algorithms are validated with ground-based measurements from Baseline Surface Radiation Network (BSRN) for the period 1992 onward. Validation for earlier periods is performed using ground data from the Global Energy Balance Archive (GEBA), and other sources. Biases relative to BSRN site measurements for the post 1992 period on all timescales are within ±5 Wm-2 with very few exceptions and root mean square differences are within 25 Wm-2 for monthly averages and 40 Wm-2 for daily averages. In general, LW errors are about half as large as corresponding SW errors.
Comparisons of global statistics of both TOA and surface fluxes and cloud radiative forcing parameters with other satellite-derived, surface-measured, and model-simulated datasets will be presented along with analyses of flux anomalies in response to many transient and interannual phenomena of the period. The entire dataset is being made available to the worldwide science community by NASA/LaRC Atmospheric Sciences Data Center at: eosweb.larc.nasa.gov/PRODOCS/srb/table_srb.html.