A synoptic climatology of short- and long-term relationships between the atmospheric circulation and river flow in the West Branch of the Susquehanna River Basin, Pennsylvania

Three linkages between the atmosphere and hydrosphere—rainfall-runon, circulation-rainfall., and circulation-runoff—for the West Branch of The Susquehanna River Basin are examined. Various spatial and temporal relationships are investigated using a circulation-environment, artificial neural network (ANN) synoptic classification. Time scales arc defined by short-term variations, ranging from one to 20 days, and long-term variations, occurring over inter-annual, time scales. Spatial domains include the large-scale, representing variations operating over an area of 106 km2, mid regional, spanning 102 − 104km2.; The ANN model is calibrated and compared to Cyeographic Information System-Hydrologic Modeling System (GIS-HMS), a distributed-parameter hydrologic model, for simulating daily spingtime (April–June) storm-flow in four nested sub- basins. ANN model calibration requires scaling river flow to fit a normal distribution, lagging precipitation by six days, and removing model bins, Model performance is highest in the larger sub-basins, where local-scale forcings are minimal. ANN model skill is affected by inter-annual variations of the Pacific/North American (PNA) and North Atlantic Oscillation (NAO) teleconnections, and the Quasi-Biennial Oscillation (QBD), which influence the spatial and temporal distribution of precipitation. In contrast to GIS-HMS, the ANN model performs better in the largest sub-basin (15,000 km2), while GIS-HMS is suited to the smaller sub-basins.; The relationship between atmospheric circulation and regional springtime (March–May) precipitation is evaluated. Five fields and four spatial domains of the atmospheric circulation are tested. A principal component (PC) filtered, 20-point domain of 850-mb geopotential heights, Jagged by one day, defines the optimal short-term coupling. This association is influenced by the PNA, NAO, and El Niño-Southern Oscillation (ENSO), which affect storm orientation and trucking over the region.; Relationships between atmospheric circulation and springtime (April–May) river flow in four nested sub-basins are examined. Five fields and two spatial domains of atmospheric circulation are tested at various temporal resolutions. Cross-spectrum analyses suggest that the strongest coupling occurs between 10 and 18 days, the time required for the passage of two or three mid-latitude cyclone's. optimal short-term relationships are represented by PC and band-pass filtered 850-mb geopotential heights (relative humidity) using 54- (20-) point atmospheric domains. Short-term associations are influenced by the PNA and ENS0, affecting the frequency of synoptic-scale systems.; These results are important for downscaling studies. First, this research has considered important methodological considerations and has demonstrated that the model is sensitive to inter-annual variations of the atmosphere. Overall, this research has shown that it is more, efficient to model river flow directly from atmospheric circulation than to model both rainfall and runoff separately.