过去300年吉林省耕地覆盖数据优化及其时空变化特征

OPTIMIZATION OF CROPLAND COVER DATA IN JILIN PROVINCE OVER THE PAST 300 YEARS AND IT’S SPATIAL-TEMPORAL CHANGE CHARATERISTICS

关键词:  可靠性评估与订正  优化  耕地覆盖  格网化分配  过去300年  吉林省

Key words:  credibility assessment and revision  optimize  cropland cover  gridded allocation  over the past 300 years  Jilin province

摘要:

目的 东北黑土区是中国重要的粮食基地，长期开垦是该地区黑土加速退化的主要原因之一，长时序黑土区耕地覆盖变化数据能为科学评价土地开垦对黑土侵蚀的影响等研究提供基础数据支撑，进而为保障黑土区粮食安全提供决策依据。方法 为了提高已有数据集的可靠性，文章通过对用于格网化分配的耕地和聚落数据原始数据的订正进而实现过去300年吉林省耕地覆盖数据的优化，并进一步进行时空特征分析。结果 （1）该研究采用的订正原始数据的方法，显著优化了格网化耕地覆盖数据结果，获得了吉林省过去300年可靠性高的格网化耕地覆盖数据集。（2）时间上看，1860年东北“封禁”政策解除为过去300年吉林省耕地变化的时间拐点，其前后耕地面积变化差异显著。（3）空间上看，过去300年间，在5′ × 5′格网尺度下，吉林省90%的格网都得到了不同程度的开垦，所有耕地格网垦殖率均为增加态势。（4）相较国际数据HYDE 3.2的结果，该研究优化后的耕地覆盖结果更好地再现了吉林省由几近荒原的新垦区逐步发展成为接近成熟的农垦区的扩张过程。（5）耕地开垦时间分布空间格局与土壤侵蚀格局不匹配，在山地丘陵地区，土壤侵蚀与开垦时长具有较好的空间对应关系，而在平原地区，土壤侵蚀与开垦时长之间的对应关系并不显著。结论 因此该研究采用的订正原始数据的优化方法，能有效避免由于原始数据缺失、遗漏等造成的格网化数据的偏差，实现显著优化格网化耕地覆盖数据结果，并且使用可靠性高的耕地覆盖数据进行相关后继研究与分析更具有科学意义。

Abstract:

The black soil region in northeastern China is an important grain base for the entire country. Long-term reclamation is one of the main causes of the accelerating degradation of the black soil in this region. Thus, the long-term cropland data in this area serve as a support for scientifically evaluating the impact of land reclamation on the black soil erosion, thereby informing decision-making to ensure food security. To optimized the cropland data in Jilin province over the 300 years, this research focused particularly on the unreliable data in the two years of 1908 and 1950. Firstly, we revised the raw data of cropland area and settlement for these years. Then, we used the settlement density method to allocate the revised cropland area to the grid scale. Finally, we analyzed the spatio-temporal characteristics of cropland changes in Jilin province over the past 300 years. The main findings were listed as follows. ① The method of correcting the raw data, adopted in this study, significantly optimized the gridded cropland cover data, and obtained a more reliable dataset for Jilin Province over the past 300 years. ② Temporally, the relaxation of "policy of prohibition", in 1860 was a critical turing point for the cropland change in Jilin Province over the past 300 years. The change of cropland area before 1860 and that after 1860 was significant difference. ③ Spatially, over the past 300 years, approximately 90% of grids at a 5′ × 5′ resolution in Jilin Province have been cultivated to some degree, with an increasing cropland fraction across all these cropland girds. ④ Compared with HYDE 3.2, the optimized cropland data in this study aligns more closely to the historical reality, effectively illustrating the cropland transition in Jilin Province from newly reclaimed to mature. ⑤ The spatio-temporal expansion of cropland did not match the pattern of soil erosion in study area. In the mountains and hills, areas with high soil erosion intensity correspond with areas having a long history of reclamation; However, in the plains, the correlation between soil erosion and reclamation duration is not significant. In conclusion, the method of raw data revision developed in this study can effectively avoid the bias in the gridded data arising from raw data loss or omission, significantly optimizing the gridded cropland cover data. The high-reliability data established in this study could be scientifically important for relevant research.