A New Empirical Estimation Scheme for Daily Net Radiation at the Ocean Surface

Ocean surface net radiation (Rn) is significant in research on the Earth’s heat balance systems, air–sea interactions, and other applications. However, there have been few studies on Rn until now. Based on radiative and meteorological measurements collected from 66 globally distributed moored buoys, it was found that Rn was dominated by downward shortwave radiation (Rg↓) when the length ratio of daytime (LRD) was greater than 0.4 but dominated by downward longwave radiation (Rl↓) for the other cases (LRD ≤ 0.4). Therefore, an empirical scheme that includes two conditional models named Case 1 (LRD > 0.4) utilizing Rg↓ as a major input and Case 2 (LRD ≤ 0.4) utilizing Rl↓ as a major input for Rn estimation was successfully developed. After validation against in situ Rn, the performance of the empirical scheme was satisfactory with an overall R2 value of 0.972, an RMSE of 9.768 Wm−2, and a bias of −0.092 Wm−2. Specifically, the accuracies of the two conditional models were also very good, with RMSEs of 9.805 and 2.824 Wm−2 and biases of −0.095 and 0.346 Wm−2 for the Case 1 and Case 2 models, respectively. However, due to the limited number of available samples, the performances of these new models were poor in coastal and high-latitude areas, and the models did not work when the LRD was too small (i.e., LRD < 0.3). Overall, the newly developed empirical scheme for Rn estimation has strong potential to be widely used in practical use because of its simple format and high accuracy.