基于植被状态指数的云南省农业干旱状况时空分析

本文首先计算了云南省2004-2013年农业干旱指数VCI,然后使用Pearson相关系数评价降水与VCI的相关性,基于VCI识别云南省2004-2013年农业干旱事件,最后,与SPEI气象干旱识别结果进行对比分析,在VCI农业干旱识别的基础上,使用干旱频率和干旱面积占比指标分析了云南省2004-2013年农业干旱时空特征。结果表明：降水只是影响VCI指数的关键因素之一;VCI和SPEI指数均能够较好对干旱进行监测并识别典型干旱,但两者的识别结果存在差异;云南省农业干旱频率在春冬两季较高,夏季较低,秋季介于夏季和春冬季之间;春夏冬三季农业干旱频率空间分布较为均匀,秋季农业干旱频率呈南低北高的分布态势,整体上北部干旱频率高于南部;2004-2013年云南省整体干旱面积占比呈现先减小后增加再波动的趋势,春冬两季整体干旱面积占比最高,分别为46.63%和47.18%,呈现下降趋势,夏季整体干旱面积占比最低,为43.81%,呈现上升趋势,秋季整体干旱面积占比介于冬春季和夏季之间,为45.74%,呈现下降趋势。总之,云南省农业干旱春冬易发性最高,影响范围最大,夏季易发性最低,影响范围最小。

Temporal and Spatial Analysis of Agricultural Drought in Yunnan Provincebased on Vegetation Condition Index

Considering the normalization characteristic of drought in Yunnan Province during recent years, it’s very significant to study the drought in Yunnan province. Nowadays, major studies focus on the meteorological drought in Yunnan Province and the studies focus on agricultural drought in Yunnan Province are few. But the data used in the study methods of meteorological drought can’t represent vegetation growth state. Also they cannot be used to evaluate the effect of drought on vegetation or analyze temporal and spatial distribution of agricultural drought in Yunnan Province. Because of these disadvantages, this paper calculated the vegetation condition index(VCI) of Yunnan Province during 2004-2013, identified agriculture drought events and analyzed temporal and spatial distribution of agricultural drought. Before the temporal and spatial distribution was analyzed, agricultural drought events identified by VCI were compared with meteorological drought events identified by the standardized precipitation evapotranspiration index (SPEI) and the Pearson correlation coefficient was calculated between VCI and precipitation to evaluate the capability of VCI index.We found the similarities and differences between these two types of drought events and analyzed the possible reasons. The results revealed that there were differences between drought events identified by these two indices because drought events identified by SPEI index are based on meteorological elements such as precipitation and temperature while drought events identified by VCI index are based on vegetation growth state which is not only affected by meteorological elements. The low Pearson correction coefficient also demonstrates precipitation is just one of the key factors which affect vegetation growth state. Though there are differences between these two types of drought events, VCI and SPEI can both monitor drought and identify classical drought events well. Based on this conclusion, temporal and spatial distribution characteristics of agricultural drought in Yunnan Province were analyzed using drought frequency index and drought-area-ration index. The results showed that: drought frequency of spring and winter is higher than that of autumn and the drought frequency of summer is the lowest. The spatial distribution of drought frequency during spring, summer and winter is relatively uniform while the drought frequency of northern Yunnan during autumn is higher than that of southern Yunnan. Overall, the drought frequency of northern Yunnan is higher than southern Yunnan and drought-area-ratio of Yunnan during 2004-2013 shows a decreasing-increasing and fluctuating trend. Drought-area-ratio index of spring and winter is the highest whose values are 46.63% and 47.18%, respectively. Both of them show a decreasing trend. Drought-area-ratio index of summer is the lowest whose value is 43.81% and shows an increasing trend. The value of drought-area-ratio index of autumn is between spring and summer and has a decreasing tendency. Based on these results, agricultural drought of spring and winter is most prone to happen and the extent is the maximum while the agricultural drought of summer is least prone to happen and the extent is the minimum. Since vegetation growth state is not only affected by drought but also can be affected by plant diseases and insect pests, irrigation, frozen injury and improper fertilization, this study monitoring agricultural drought based on VCI index has some limitations. Future work should focus on the physical mechanism of agricultural drought and the biophysical response of vegetation to drought in order to monitor and forecast agricultural drought more accurately.