基于动态过程导向的马铃薯种植适宜性时空精细化评价研究

SPATIO-TEMPORAL REFINED ASSESSMENT OF POTATO PLANTING SUITABILITY BASED ON DYNAMIC PROCESS ORIENTATION

关键词:  马铃薯  适宜性  精细评价  过程导向时空加权  随机森林

Key words:  potato  suitability  refined assessment  process-oriented  random-forest

摘要:

目的 为优化我国冷凉区作物种植结构布局，选择我国东北黑龙江、吉林和内蒙古冷凉地区为研究区，基于动态过程导向的马铃薯种植适宜性时空精细化评价方法，开展典型马铃薯早熟、中晚熟品种种植适宜性研究。方法 文章根据马铃薯干重与生育时间函数关系获取以每天为评价基本单元的纵向加权精细化赋值，基于随机森林方法获取横向环境指标精细化加权值，生成马铃薯种植适宜性评价结果图；应用马铃薯空间实际种植点及DSSAT-SUBSTOR-Potato模型模拟马铃薯单产，对适宜性评价结果进行交叉验证。结果 （1）早熟品种费乌瑞它的种植适宜区集中于大兴安岭林区与平原交界的缓冲地带，而马铃薯中晚熟品种延薯四号适宜区分布在平原、林区与平原交界处。无论是费乌瑞它或是延薯四号，其在牡丹江林前耕地区域的适宜性普遍不高，但是在牡丹江周围的平原耕地适宜性较好。（2）基于动态过程导向的马铃薯种植适宜性时空精细化评价方法与模型评价方法有关早熟品种的评价结果较为一致，重合度超过70%；在生长期较长的中晚熟品种的适宜区评价中，由于权重和生长期差异导致评价结果出现不同，尤其在低适宜区差异更加明显，重合度低于40%。结论 由此可知，应用不同方法评价作物种植适宜性时，不同的品种可能获取不同的结论。

Abstract:

In order to optimize crops planting layout in the low-temperature region of China, this study selects the low-temperature zones in the provinces of Heilongjiang, Jilin and Inner Mongolia as the study areas, and adopts a refined evaluation method for potato planting suitability based on the dynamic process-oriented, which is for realization of assessment indicators to be more precise and accurate in the spatio-temporal dimension, then conducts a study on the suitability of planting typical potato varieties of early, middle and late maturing.In terms of the function relationship between dry weight values of potato and crop growth thermal time, the different potato growth phase could be weighted on a daily basis, and meanwhile biophysical indicators were weighted by using the random forest method. Moreover, the maps of potato planting suitability assessments were available. The actual potato planting spots and simulated results of the DSSAT-SUBSTOR-Potato model were used to cross-check the suitability assessment results. The Results were listed as follows. (1) The suitably planting areas of the early maturing variety, the Favorita, concentrated on the fringe areas between forest and plain in the Greater Khingan Mountains, while the suitable planting areas of middle and late maturing variety Yanshu No. 4 mainly distributed in the plain and the interacting regions between plain and forest. Whether Favorita or Yanshu 4, suitability in the farmland neighbor to forest was unoptimistic, but it was really high in the plain farmland around the Mudan River. (2) The assessment results for suitability from above-mentioned method and DSSAT-SUBSTOR model coincided with each other with a degree more than 70%. However, in the suitability evaluation for middle-late maturing varieties, results differentiate a lot, especially in the low suitability areas with a consistence degree lower than 40%.Conclusion It is concluded that results are different with application of assessment methodologies and targeting at different varieties.