Integrated effects of rainfall regime and canopy structure on interception loss: A comparative modelling analysis for an artificial larch forest

Model optimization is essential for accurately simulating rainfall interception loss (I) in artificial forests and for quantifying the influence of canopy structure, rainfall regime, and meteorological variables on I. Herein, the revised Gash (addressing physical mechanisms) and Wang (addressing canopy structural impacts) models were used for comparative I simulations using a larch plantation in NW China. We used throughfall (T-f) and stemflow (S-f) measurements from the growing seasons of three consecutive years and found that from the total rainfall measured (1371.6 mm) in 59 selected rain events, T-f, S-f, and I accounted for 80.8%, 1.6%, and 17.6%, respectively. Surprisingly, the two tested models showed similar behaviours in their I simulations, with their performances classified as "very good," exhibiting 5% and 2.3% cumulative I relative errors for the revised Gash and Wang models, respectively. We noted that, although I was well simulated for light-moderate rain events (<25 mm), absolute errors increased significantly as rainfall amount and duration increased, owing to the constant evaporation rate applied in the models. Leaf area index (LAI) explained 40% of the water detention variation in the canopy, rather than I, which only reflected canopy storage capacity dynamics during the growing season. Collectively, our results revealed that models only addressing the independent impact of wet evaporation or structural dynamics are inadequate for accurately simulating I during heavy rain events (>25 mm). Therefore, future models should account for the potential errors arising from evaporation estimation and consider the effects of LAI.