Probing the Error Term in Streamflow Reconstruction

Water balance models were first applied indirectly by tree-ring researchers in reconstructions of Palmer Drought Severity Index in the mid-1970s. Recent years have seen more direct and explicit use of water-balance models. This talk explores how a simple monthly water-balance model can be applied to diagnose possible causes of large errors in reconstructions of streamflow. The setting for the case study is the Chemora River Basin, a small (<800 km2.) watershed in northeastern Algeria with nearby climate-sensitive tree-ring chronologies and short records of precipitation and gaged flow.