关键词: 大数据, 时空尺度, 分布式生态模型, 生态系统服务流, 非生物流

Abstract: Forsythia suspense (Thunb.) Vahl. is a kind of bulk medicinal material. It is also a common species for ornamental use and for water and soil conservation. Based on 119 current geographical distribution records of F. suspense and 12 related factors, including climate, soil and topography, the potential geographic distribution of F. suspense was assessed by generalized linear models (GLM), random forest (RF), generalized boosting model or usually called boosted regression trees (GBM/BRT), and artificial neural network (ANN) in the BioMod2 package. The results showed that the four models had good performance, among which the RF model had the best efficiency, with high values of evaluation indexes, including AUC (0.95), TSS (0.81), and KAPPA (0.68). For the growth of F. Suspense, the sensitivity of temperature was higher than that of precipitation. The potential geographic distribution of F. suspense situated in the mountains and along rivers in Shaanxi, Shanxi, Hubei, Henan, Liaoning. The suitable habitats of F. suspense are characterized by an altitude of 250-2200 m, annual mean temperature of 12.1-17.3 ℃ and annual precipitation of 400-1000 mm. This study provides guidance for F. suspense’s planting.

Key words: distributed ecosystem model, big data, ecosystem service flow, abiotic flow., spatial and temporal scale