A spatial simulation model for karst rocky desertification combining top‐down and bottom‐up approaches

• Published in: Land degradation & development Vol. 29; no. 10; pp. 3390 - 3404
• Main Authors: Xu, Erqi, Zhang, Hongqi
• Format: Journal Article
• Language: English
• Published: Chichester Wiley Subscription Services, Inc 01-10-2018
• Subjects: Computer simulation; Contraction; Desertification; desertification restoration; dynamics simulation model; Evolution; Information processing; Iterative methods; Karst; karst rocky desertification; Model testing; Modules; Restoration strategies; Reversion; scenario setting; spatial heterogeneity
• ISSN: 1085-3278; 1099-145X
• DOI: 10.1002/ldr.3103

Karst rocky desertification (KRD) is a serious ecological and environmental issue threatening southwestern China. The evolution of KRD calls for predicting its potential expansion or contraction, to support policy‐making for combating KRD. A spatial model named ‘SDKRD’ for simulating KRD dynamics was developed in this study. The novelty of it is the ability to explore spatially local KRD evolution and integrate this information into a global estimation. The SDKRD model consists of three modules: a global KRD forecast and scenario setting, a local exploration of spatially varying effects of driving forces and neighbourhood influence on KRD evolution, and an iteration module for combining results of aforementioned modules. The SDKRD model was tested in the Changshun County as a case‐study. The accuracy assessment confirmed a spatially visual consistency and a robust statistical result. Spatial KRD maps from 2010 to 2030 were simulated under three scenarios. The historical extrapolation scenario found a spatially global KRD reversion and locally scattered KRD deterioration. Different KRD restoration strategies under the other two scenarios could reverse the transformation of KRD classes at different magnitudes. Intermediate classes in the KRD evolution process could be more easily reversed. Under three scenarios with the increased strength of KRD management measures, areas of no KRD increased by 8.7%, 13.5%, and 16.6%, and areas of moderate KRD decreased by 38.3%, 53.0%, and 53.9%, respectively. SDKRD model presents specific KRD evolution trajectories and visualizes the KRD transformation at each location, which can help find cost‐effective ways of combating KRD with differentiated strategies.