Long-Lead Streamflow Forecasting in the Southwest of Iran by Sea Surface Temperature of the Mediterranean Sea

Abstract

Studying the relationship of ocean/atmospheric modes with hydrologic cycle in lands would be helpful in hydrologic and meteorological long-lead forecasting, climate fluctuation awareness, and water resources management. One way of measuring these phenomena is to use sea surface temperature (SST) which has received so much attention lately. Local experiments are highly recommended for evaluating effective regions of the seas in correlation with the hydrological responses of the study areas. In this paper, the singular value decomposition (SVD) technique is applied for developing temporal expansion series of Mediterranean SSTs according to the precipitation and streamflow variability of southwest of Iran. These parameters are then used in a nonparametric model, which is modified in a way to deal with the experienced forecasting errors. A lead-time approach was adopted, such that the autumn (October–December) SSTs were correlated with winter (January–March) precipitation and February–May averaged streamflow values. The application of SSTs resulted in better forecast for all stations in dry years rather than wet years. The results showed better capability of SSTs temporal expansion series (first mode) than those of well-known climate indices [Southern Oscillation Index, Pacific Decadal Oscillation (PDO), North Atlantic Oscillation (NAO), and Atlantic Multidecadal Oscillation (AMO)] to be used as the predictors of streamflow in the case study of this paper.