浙江省植被物候变化及其对气候变化的响应

基于改进的Savitzky-Golay滤波算法重构了2001-2010年MODIS NDVI时间序列数据,反演了浙江近10 a来的植被生长状况,利用分段多项式拟合和动态阈值法提取了自然植被重要物候期(生长季开始时间、结束时间和生长季长度等),分析了植被物候期的年际变化趋势和空间分异特征,并结合同期气象数据,探讨了植被物候变化与气候变化的关系.主要结论如下:①浙江植被覆盖有所减少,整体下降趋势不显著,平均植被生长季为222 d,长度略有延长,其中开始时间提前趋势不显著,而结束时间推迟显著.②植被生长季为3月下旬至11月中旬,植被生长季的开始时间从北往南逐渐推迟,生长季结束时间相对集中,其中生长季延长的区域面积和缩短的区域面积相当.③气象要素与关键物候期参数的相关性分析表明冬季的热量供给是影响浙江植被生长开始的重要因子,植被生长季前期温度积累的增加有利于植被生长;生长季结束时间与当年温度呈极显著正相关,与当年降水和湿润指数的减少呈负相关,但影响不显著,从与各相关月、季度相关分析来看,秋季干湿程度对植被生长季结束时间影响相对较大.

Vegetation Phenological Variation and Its Response to Climate Changes in Zhejiang Province

In order to understand the vegetation phenological variation of Zhejiang Province and further improve the vegetation response to climate change mechanism, the vegetation phonological variation and relationship with the climate factors were deeply analyzed based on the reconstructed MODIS NDVI time-series using the improved Savitzky-Golay algorithm, the vegetation growth condition was retrieved, and the vegetation phenological metrics (the beginning data, the ending data and length of growing season) were extracted adopting the piecewise polynomial fitting and dynamic threshold method, the vegetation phenology interannual variation trend and its spatial pattern were analyzed, the relationship between the vegetation changes and the climate variations were discussed combined with the meteorological data. The results showed that the vegetation cover expressed slowly decreasing trend, while the length of growth season was slightly extended. The vegetation growing season was from late March to mid November, and the average length of growing season was about 222 days. In the past 10 years, the beginning time of growth season expressed the advance trend, but not significant, and the ending time was significantly delayed. The length of growing season was extended, about 7 days every 10 years. The spatial pattern of the begining time of growing season was delayed from north to south, the ending time was relatively concentrated in November, vegetation growing season of extended area was similar to the shortened area, about 32.40% and 31.88% of the effective pixel respectively. The correlation analysis of phenological metrics and climate factors indicated that heat supply in winter is the important factor for vegetation growth, and the increased cumulative temperature could prompt the start of vegetation growing season. The end of the growing season was positively related with mean temperature while precipitation and humid index were negatively correlated, but not significant, from the view of the month and seasonal scale, the humidity in autumn affected larger on the ending time of growing season.