基于MODIS的重庆市植被指数对地形的响应

Response of Vegetation Index Based on MODIS to Topographic Factors in Chongqing

利用重庆市2011年3-11月晴空条件下逐日MODIS数据,采用最大值合成法合成逐旬(月、年)归一化植被指数,将各旬植被指数按高度和坡度分别分成6个高度特征区和5个坡度特征区,对各特征区植被指数进行对比分析。结果表明,(1)区域平均植被指数随特征区高度或坡度值增加呈增大的趋势。高程1500m以下相邻区域间NDVI值相差3.5～8.7,且植被指数随高度升高增加较快,1500m以上相邻区域间NDVI值相差在1.9以内,随高度升高增加缓慢;坡度25°以下相邻区域间NDVI值相差3.0～8.0,且植被指数随坡度增加增大较快,25°以上相邻区域间NDVI相差0.2～1.0,随坡度增大增加缓慢。(2)大于平均植被指数的频率与高度和坡度的相关分析表明,在海拔较高的特征区,植被指数高值格点的频率也较大,该频率与海拔高度间呈极显著正相关关系,相关系数为0.9552(P<0.01);而植被指数与坡度无显著的相关性,相关系数为-0.3000。(3)按高度分区比按坡度分区更能体现植被指数的地区差异性。各高度特征区高程与植被指数的相关系数在0.9541～0.9871(P<0.001),各坡度特征区坡度与植被指数的相关系数在0.9457～0.9723(P<0.01)。研究结果说明,重庆市地形因素对植被指数有较大的影响,且高度因素对植被指数及其分布的影响较坡度更明显。

Based on the daily MODIS data under clear sky condition from March to November 2011 in Chongqing, authors computed the 10- day （monthly,yearly） NDVI by using MVC method. The serial NDVI enlarged 100 times in different period of time were divided into 6 characteristic districts with different altitudes which were below 400m,400 -800m,800 -1200m, 1200 -1500m, 1500 -2000m and above 2000m and 5 characteristic districts with different slopes which were below 5°,5° - 15°, 15° - 25°,25°- 35° and above 35°, and the NDVI of each district was analyzed . The results showed that （ 1 ） The NDVI augmented with the increase of altitude or lope. NDVI in districts below 1500m increased conspicuously and the added value was 3. 5 to 8.7 between neighboring districts, but in districts above 1500m increased inconspicuously and the added value was 〈 1.9 between neighboring districts. NDVI in districts below 25° increased conspicuously and the added value was 3.0 to 8.0, but in the districts above 25° increased inconspicuously and the added value was 0. 2 to 1.0. （2） According to analysis,in higher districts, the grid frequency of NDVI above the average augmented with the increase of altitude. The correlation coefficient between NDVI and altitude was 0. 9552 （ P 〈 0. 01 ） while that between NDVI and slope was - 0. 3000. The results showed that the grid frequency of NDVI a- bove the average augmented with increase of altitude while there was no significant correlation between grid frequency of NDVI above the average and slope. （ 3 ） The difference of NDVI in different districts was reflected more conspicuous- ly by altitude division than by slope division. The correlation coefficient was more than 0. 9541 between NDVI and alti- tude （ P 〈 0. 001 ）, and was 0. 9457 -0. 9723 between NDVI and slope （ P 〈 0. 01 ）. All the results above showed that in Chongqing city, the topographic factors had obvious effect on NDVI, and the effect of altitude on NDVI was more con- spicuously than that of slope.