The modelling of rainfall interception in growing and dormant seasons for a pine plantation and a black locust plantation in semi-arid Northwest China

Abstract

Interception loss can remove a significant portion of rainwater from forested ecosystems. Therefore, the quantification and modelling of interception loss are of significant importance if human and ecosystem water demands are to be balanced under a future changing climate. This is particularly true for semi-arid/arid regions, where afforestation has become an important ecological restoration measure to tackle desertification, poverty and climate change. However, quantification and modelling of interception loss of plantations in these regions have rarely been reported. In the present study, rainfall interception loss was quantified and modelled over a one-year period (January-December 2016) for a deciduous broad-leafed R. pseudoacacia plantation and an evergreen needle-leaf P. tabuliformis plantation (common afforestation tree species) situated in the semi-arid Loess Plateau of China. The stand age, density, canopy cover and leaf area index of R. pseudoacacia during the study period were 15 years, 2000 tree ha−1, 0.48 and 1.41 m2 m−2, respectively. The corresponding values for Pinus tabuliformis were 17 years, 1200 tree ha−1, 0.62 and 2.53 m2 m−2. The measured throughfall, stemflow and derived estimates of interception loss for R. pseudoacacia were 81.1%, 1.3% and 17.6%, respectively. The corresponding values for P. tabuliformis were 75.4%, 0.7% and 23.9%. Given that the weather conditions experienced by the two forest stands were similar, the observed higher interception loss for P. tabuliformis can be explained by the higher canopy storage capacity and wet canopy evaporation rate of this species. The revised Gash analytical model of rainfall interception was well calibrated and validated against field measurements and was able to simulate the cumulative interception loss at two forest stands accurately, and it also effectively captured the seasonal variation (leafed growing and leafless dormant seasons), provided that the optimized wet-canopy evaporation rates were used. The revised model was highly sensitive to the canopy storage capacity and changes in the ratio of mean wet canopy evaporation to mean rainfall intensity and less sensitive to canopy cover, but it was found to be fairly insensitive to the trunk storage capacity.