新疆棉花物候时空变化遥感监测及气温影响分析

Monitoring Spatial-temporal Change of Cotton Phenology in Xinjiang and Its Response to Climate Change

基于2001～2016连续16a的MOD09Q1数据计算获取NDVI时序数据,利用Savitzky-Golay(S-G)滤波重构NDVI时序曲线,以动态阈值法提取新疆大地块棉花的生长季开始期、生长季结束期和生长季长度信息,并分析新疆棉花物候时空变化特征及其对气温变化的响应。结果表明:(1)南北疆棉花物候空间差异显著:生长季开始期由南向北逐渐推迟,南疆集中于第151d之前,北疆于第151～163d之间;生长季结束期逐渐提前,北疆大部分于第292d前结束生长,而南疆集中发生在第298d之后;生长季长度逐渐缩短,南疆普遍长于150d,北疆通常短于150d。(2)南北疆物候变化趋势不一致:阿克苏生长季开始期和结束期表现为"推迟—提前—推迟—提前"变化趋势,生长季长度不明显延长;沙湾县生长季开始期表现为微弱提前趋势,生长季结束期则为先推迟再提前趋势,生长季长度表现为缩短、延长、缩短的波动变化。(3)原因分析表明棉花物候主要受气温影响:生长季开始期与10℃初日、15℃初日、生长初期平均气温表现显著的相关性;生长季结束期与10℃终止日相关性较好。

Observations of vegetation phenology provide valuable information regarding ecosystem response to environmental conditions, especially to climate change. Cotton is one of the most important economic crops in Xinjiang,and its phenological change can directly reflect the change of climate in Xinjiang.This re- search was an attempt to extract cotton phenological parameters in Xinjiang by using 16 years＇ （2001 to 2016） time series MODIS Normalized Difference Vegetation Index（NDVI）：firstly,filtering noise from the time series data using Savitzky-Oolay filtered method; then detecting cotton phenology parameters （Start of Growth Season（SOS）,End of Growth Season（EOS）,Long of Growth Season（LOS）） using Dynamic Threshold method;finally, the spatial patterns and temporal trends of observed cotton phenological charac- teristics were analyzed over the past 16 years and the relationship between cotton phenology and tempera- ture changes was also discussed.The result of this study showed that the spatial patterns of cotton phenolo- gy were significantly different in study region： SOS delayed gradually from Nanjiang to Beijiang, and mainly occurred before 151^st and after 151^st days respectively;EOS gradually advanced,most areas of northern Xinjiang ended up 292^nd days ago,while the southern Xinjiang happened 298th days later;LOS shortened,Nanjiang is generally longer than 150 days while Beijiang is usually shorter than 150 days.The trend of cotton phenology（2001-2016） under climate change in northern and southern Xinjiang were not completely similar：SOS and EOS in southern Xinjiang showed a delay-advancing-delay-advancing trend,and LOS was un- significantly delayed;While SOS in northern Xinjiang were slightly advanced and EOS exhibited a delay trend followed by an advancing, LOS showed a shorten-lengthen-shorten trend.In addition, cotton phenology showed a strong correlation with the temperature：SOS and EOS were positively correlated with the beginning date of 15℃ and the end date of 10℃ respectively;SOS was negatively correlated with the spring temperature,while EOS had a positive correlation with autumn temperature.