Generating a High-Resolution Time-Series Ocean Surface Net Radiation Product by Downscaling J-OFURO3

Abstract

The ocean surface net radiation ( R-{n} ) characterizing ocean surface radiation budget is a key variable in ocean climate modeling and analysis. In this study, a downscaling scheme was developed to generate a high-resolution (0.05°) time-series (2002-2013) daily ocean surface R-{n} from the third-generation Japanese Ocean Flux Data Sets with Use of Remote-Sensing Observations (J-OFURO3) at 0.25° based on the Advanced Very-High-Resolution Radiometer (AVHRR) top-of-atmosphere (TOA) observations (AVH021C) and other ancillary information (Clearness Index and cloud mask). This downscaling scheme includes the statistical downscaling models and the residual correction post-processing. A series of angle-dependent downscaling statistical models were established between the daily ocean surface R-{n} in J-OFURO3 and the AVHRR TOA data, and then, the residual correction was conducted to the model estimates R-{n} _AVHRR_est to obtain the final downscaled data set R-{n} _AVHRR. Validation against the measurements from 57 moored buoy sites in six ocean observing networks shows the high accuracy of the downscaled estimates R-{n} _AVHRR_est with a R^{2} of 0.88, RMSE of 23.44 text{W}cdot text{m}^{-2} , and bias of -0.14,,text{W}cdot text{m}^{-2} under all-sky condition. The results of the spatio-temporal analysis in R-{n} _AVHRR and intercomparison with Cloud and the Earth's Radiant Energy System (CERES) and the European Centre for Medium-Range Weather Forecasts (ECMWF) Interim Re-Analysis (ERA-Interim) products also indicated that the superior of the R-{n} _AVHRR with more detailed information especially in the hot spot regions, such as central tropical Pacific (warming pool), Atlantic and Equatorial Eastern Indian Ocean (EIO).