科研之友
微信
新浪微博
Facebook
分享链接
摘要
A wide range of scenario studies aiming at rural development require regional patterns. of crop yield. This study aims to evaluate three different modeling approaches for their suitability to assess regional potato yield patterns. The three model approaches include (1) an empirical model; (2) a process-based crop growth simulation model; and (3) a metamodel derived from the crop growth simulation model. Scenario studies have specific requirements for these modeling approaches including (1) their ease to use, (2) a realistic sensitivity, (3) the relevance in terms of generating the desired system property, and (4) their credibility in producing recognizable plausible outputs for stakeholders. The modeling approaches were applied to assess patterns of potato yields in a major production area in northern Ecuador. All three modeling approaches require significant expert knowledge for their development and calibration. However, after this initial phase, the empirical model and the metamodel are very easy to use and transparent. However, their application domain is limited to the case study area. The application of the crop growth simulation model remains complex and the model functions as a black box. The results show that regional patterns of potato yield are determined by a limited number of variables. The sensitivity of all three modeling approaches to climatic factors and water holding capacity suggest that the potato production in the area is constrained by water availability and temperature. All models generate similar yield patterns. However, the empirical model derives quality adjusted potato yields that correlate highly to the observed yields, whereas the crop growth simulation model and the derived metamodel produce potential, water and nutrient limited yields. Scenario studies may require yield patterns at different levels of resolution. All results could be aggregated to different resolutions but in general the patterns remained very similar. All three modeling approaches were capable to reproduce the observed regional pattern of potato yield and are therefore considered to be credible. In analyzing the effect of spatial aggregation on the performance of the modeling approaches, the results show that aggregation improves the overall correspondence between model output and interpolated, observed yields. It can be concluded that the various modeling approaches have their unique value. They are therefore complementary to each other for the interpretation of the observed patterns. The patterns themselves do not vary much and as such the most convenient modeling approach can be selected (based on available expertise and data).
- 出版日期2013-7
