1982-2013年潞安矿区NDVI3g变化趋势及气候响应

为了揭示全球变化背景下矿业开发活动对矿区生态环境的影响，采用长时序GIMMS AVHRR NDVI3g（1982-2013年）全球植被指数数据集和气候信息（年均降水量和气温数据），运用IDL编程实现数据合成运算、线性回归和趋势拟合，从时间、空间、气候三方面对矿区、缓冲区（10 km、20 km）、校验区的NDVI平均值和总值进行比较研究，并推算出生长期变化趋势.时序分析表明，32 a来矿区NDVI总量随开采年限延长呈先增后减/波动中下降的趋势，下降速率为0.18/（10 a）.矿区植被返青期滞后3 d，枯黄期提前30 d，生长期缩短33 d，缩减速率为10.3 d/（10 a）.空间分析表明，除潞安矿区外其他三区生长季均有所延长，研究区平均NDVI年增长率依次为矿区（1.09%） < 10 km缓冲区（2.16%） < 20 km缓冲区（8.86%） < 校验区（9.87%），矿区NDVI总量自1995年后开始减少，非开采区NDVI总量呈增加趋势.气候变化分析表明，校验区NDVI对温度敏感性高于降水量，矿区NDVI对降水量敏感性高于温度，其中温度对两区植被生长有明显滞后性.研究显示，矿业开发活动抑制了矿区及周边区域NDVI的增长，NDVI年增长率远低于校验区，受温度升高、降水量减少共同作用，自然生态下校验区NDVI呈增加趋势，生长季延长；而受开采扰动影响下的矿区植被活动呈减弱趋势，生长期也有所缩短. 

Variation Trend and Climate Response of NDVI3g in Lu'an Mining Area from 1982 to 2013

Research on long time series Normalized Difference Vegetation Index (NDVI) dynamic changes in mining areas is beneficial to clarify the impacts from global change and to reveal the influence of human activities on the mining area ecological environment. The long time series GIMMS AVHRRNDVI3g (1982-2013, 32 a) global vegetation index and climate information (i.e., precipitation and temperature) were selected as data sets. Data synthesis operation, linear regression and trend fitting contrasting the maximum NDVI, average NDVI and the total NDVI of directly affected areas, buffer areas (10 km and 20 km) and checked areas were used to calculate the change trends of the start, end and length of growing season. Three components in time, space and climate were implied using IDL programming language. Analysis of the sequential correlation results indicated that due to global warming, the mining area start of growing season (SOS) has been postponed 3 d, end of growing season (EOS) has advanced 30 d and length of growing season (LOS) has shorten 33 d -a reduction rate of 10.3 d/(10 a). With the increasing of mining years, the annual average of total NDVI declined during the 32 year research period, with a decline rate of 0.18/(10 a). Analysis of the spatial correlation indicated that the LOS in the buffer area (10 km and 20 km) and checked area (CK) have been extended, and the LOS in the mining area has been shortened. The LOS in mining area is 3 days shorter than that in the CK. The annual average NDVI over 32 years was 0.2936, 0.2964, 0.3250 and 0.2918 for the mining area, the 10 km buffer area, the 20 km buffer area and the CK respectively. The relationships of different research districts were as follows:annual NDVI growth rate was followed by mining area (1.09%) < 10 km buffer area (2.16%) < 20 km buffer area (8.86%) < CK (9.87%). The annual NDVI growth rate in the mining area was lower than that in the CK. The total NDVI in the mining area began to decrease since 1995, while the total NDVI in non-mining area showed an increasing trend. This indicated that coal mining has had an obvious effect on regional ecological processes. The analysis of climate change correlation indicated that the NDVI is increasing with the increasing of temperature and decreasing of precipitation. The annual NDVI growth rate in natural ecological CK is higher than that in the mining area; the LOS in the CK showed a trend of extended with increasing of temperature and decreasing of precipitation, while the LOS in the mining area showed a shortening trend. The NDVI in the mining area, in addition to the interference by climate change, is also the result of mining activities. The results show that it is an indisputable fact that mining has a significant impact on regional ecological processes. (1) Human activities interfere with the evolution process of natural ecology over mining area. (2) Climate change has caused the NDVI to increase over the whole area. (3) Mining activities lead to decreasing NDVI in both mining and non-mining area. (4) The NDVI is influenced more seriously by mining activities than climate change.