PP51A-2265
Diagnostic Value of the Errors in Dendrohydrologic Reconstruction
Friday, 18 December 2015
Poster Hall (Moscone South)
David M Meko1, Ramzi Touchan1, Dalila Kherchouche2 and Said Slimani3, (1)University of Arizona, Tucson, AZ, United States, (2)The University Hadj-Lakhdar, Institute of Veterinary and Agronomy Sciences, Batna, Algeria, (3)The University Mouloud Mammeri, Faculty of Biological Sciences and Agronomy, Tizi Ouzou, Algeria
Abstract:
Regression based dendrohydrologic reconstructions of streamflow yield valuable information on hydroclimatic variability, especially where gauged flow records are short. Yet even the most accurate reconstructions seldom explain more than 2/3 of the variance of annual flows. A reconstruction of the Chemora River, northeastern Algeria, is generated from a tree-ring network of 23 chronologies of Cedrus atlantica and Pinus halepensis, and the error term is examined for information on possible factors contributing to large misses in reconstructed flows. The reconstruction in this example explains more than 60% of the variance of annual flows over a 1971-2002 calibration period, and provides reconstructed flows to 1853. Reconstruction errors from the two-species reconstruction are evaluated for possible relationship to form of the annual hydrograph and to derived seasonal variables (e.g., snow water equivalent, runoff, soil moisture) from a monthly resolution water-balance model.
