Modelling canopy interception loss from a Madrean pine-oak stand, Northeastern Mexico

Throughfall associated with 34 rainfall events was measured from Year Day 1, 1999 to Year Day 184, 2001 within a pineoak stand of the central Sierra Madre Oriental mountain range of northeastern Mexico. Throughfall and canopy interception loss accounted for 83.6 +/- 2. 1 % and 15.8 +/- 1.8% (significance level, alpha, = 0.05) of the 691.8 turn cumulative rainfall input, respectively. A reformulated Gash analytical interception loss model and the Liu analytical model run with predetermined model parameters derived from the literature underestimated cumulative interception loss by 34.4%, and 37.1%, respectively. When the models were run with study period derived parameters the reformulated Gash model overestimated the observed interception by 3.0%, while the Liu model underestimated observed cumulative interception by 0.8%. Although good agreement between observed and estimated cumulative canopy interception loss was found using the analytical models with study period derived parameters, relatively poor agreement was found at the rainfall event scale. Although the analytical versions of the reformulated Gash and Liu models provided similar results, the Liu model is recommended for further application in the study area since it requires less data input and follows an exponential wetting of the canopy approach, something that the data collected during the study supports. Future work regarding the Liu model is discussed. Copyright (C) 2007 John Wiley & Sons, Ltd.