退耕还林（草）工程实施前后黄土高原植被覆盖时空变化分析

基于GIMMS NDVI3g（the third generation of Global Inventory Modeling and Mapping Studies Normalized Difference Vegetation Index）数据,辅以趋势分析、Mann-Kendall检验、Hurst指数等方法,识别了1982—2013年及1982—1999、2000—2013年黄土高原植被覆盖时空演变特征,并探讨其驱动因素。研究发现：1）1982—2013年及1982—1999、2000—2013年期间黄土高原生长季NDVI分别以0.019/10 a（P<0.01）、0.016/10 a（P<0.05）和0.057/10 a（P<0.001）的速率增加;2）除1999年以前林地外,所有植被类型的生长季NDVI均呈现显著的增加趋势,2000—2013年尤为明显;3）黄土高原生长季NDVI呈现由东南向西北递减的趋势,1982—2013年及1982—1999、2000—2013年NDVI显著上升的面积分别占74.94%、24.26%和53.34%,主要集中在黄土高原的北部和中部地区;4）研究区未来生长季NDVI呈持续性和反持续的比重分别为33.32%和66.68%,其中持续改善和由改善变为退化的面积分别占31.08%和61.88%;5）2000年以后降水增多与生长季NDVI上升相对应,大规模的生态工程建设对2000—2013年生长季NDVI增加有重要影响。

Spatiotemporal Variation of Vegetation Coverage before and after Implementation of Grain for Green Project in the Loess Plateau

As an ecologically vulnerable area in China, the Loess Plateau has been indicated by many studies to have a significant increase trend in vegetation coverage. However, it is required to clarify the growing trend before and after the implementation of the Grain for Green Project. Based on the third generation of Global Inventory Modeling and Mapping Studies Normalized Difference Vegetation Index (GIMMS NDVI3g) dataset, this study investigated the patterns of spatiotemporal variation of vegetation coverage and the possible environmental factors in the Loess Plateau during 1982-2013, 1982-1999 and 2000-2013 using the Sen+Mann-Kendall model, partial correlation analysis, Mann-Kendall test and Hurst index. The results are as follows: 1) the growing season Normalized Difference Vegetation Index (NDVI) of the study area increased progressively during the periods of 1982-2013, 1982-1999 and 2000-2013, with the rate of 0.019/10 a, 0.016/10 a and 0.057/10 a, respectively. 2) The growing season NDVI of five vegetation types significantly increased during the three periods, especially during 2000-2013, except the forest during 1982-1999. 3) Vegetation coverage showed a decreasing trend from southeast to northwest, and there was an abrupt decrease in the middle of the Loess Plateau. 4) The reverse characteristics of growing season NDVI change were stronger than the same characteristics in the Loess Plateau. Some 31.08% of the Loess Plateau will follow a continuous increasing trend, while 61.88% of the entire study area is predicted to decrease in future. 5) The vegetation coverage was mainly affected by climate change and human activities. Our results also indicated that increasing precipitation is a factor that results in the improvement of the vegetation coverage. In terms of human activity, our result indicted that there was a strong correlation between the cumulative afforestation area and growing season NDVI during 1999-2013.