| Drought monitoring on the regional ecosystems is becoming more and more imporant with the development of social economic development and space technology.In recent decades,the occurrence of drought events has shown a trend of accerelation as the pace of global climate change.Studies on drought monitoring have made great progress in recent years.A number of methods especially the index approaches indicating drought intensity have been developed for practical applications such as in agriculture.It can be said that the available methods for drought monitoring can be divided into two categories:the meteorological index approaches that based on meteorological observation data and the remote sensing index methods that based on remote sensing observation data.Moreover,since agriculture is the economic sector heavily affected by drought events,many available drought monitoring approaches are established for agricultural applications,which consequently results in the agro-drought monitoring being the most important scope of drought researches.With the economic and social developments of China,construction of ecological civilization has been greatly strengthened.Given the basis of agricultural drought monitoring,a further expansion of the global ecological drought monitoring has become a vital development direction of drought monitoring.Hunan Province is located in the subtropical hilly and mountainous areas,dominated by red with a large forest cover,and is a typical ecological province.At present,there are few studies on ecological drought monitoring in Hunan Province.It is of great importance to carry out in-depth monitoring of ecological drought in Hunan Province and propose suitable ecological drought monitoring techniques.Long time-series meteorological data and MODIS remote sensing data were used in the study to analyze the spatial-temporal process of ecological drought in Hunan Province.Two drougt indices were computed from the data:the drought index from meteorological data and the index from remote sense data.Spatial-temporal variation of ecological drought in the province were examined on the basis of the two drought indices.Since the province is a typical red soil hilly region in Southern China,the study on ecological drought will be very importance.It can provide some useful technological helps for future implementation of ecological drought monitoring in Southern China Region.The meteorological monthly data from 1961 to 2013 were used to compute Standardized Precipitation Evapotranspiration Index(SPEI)as the drought index or indicator of ecological drought in Hunan province.On the basis of the index,we analyzed the inter-annual and monthly changes of ecological drought in the province.The spatial distribution characteristics of different time scales in each ecological division were revealed.In the aspect of remote sensing index,the MODIS data is mainly used for the study.First,we standardize the vegetation supply index through improving the avalable Vegetation Supply Water Index(VSWI)model according to the magnitude of vegetation index in each pixel.Then we integrated the VSWI with the comprehensive precipitation anomaly index to build an ecological drought comprehensive index,which we termed as Ecological Drought Comprehensive Index(EDCI).Finally we used the EDCI from remote sensing data to analyze and reveal the temporal and spatial variation characteristics of ecological drought in Hunan Province.The improved CASA(Carnegie-Ames-Stanford approach)model was used to validate the monitoring efficiency of EDCI with its estimation of the net primary productivity of vegetation and response of the change of NPP of vegetation to ecological drought in the province.The main conclusions of our paper are as follows:(1)The SPEI index-based ecological drought monitoring method mainly adopted the Man-Kendall method and wavelet analysis of time series,as well as the ENSO events and their indices.The results showed that SPEI within multiple time scales can better reflect the drought situation of Hunan Province in terms of time variations,and the spatial SPEI index can better reveal the spatial distribution characteristics of ecological drought in Hunan Province.On the monthly scale,the values of Hunan SPEI-1 and ENSO cold and warm events are counted,ENSO cold and warm events,espeically ENSO cold events,are more likely to cause drought.On the annual scale,the cross-wavelet transform analysis of SPEI-12 and MEI indicate that SPEI-12 has an oscillation period of 1.0?3.5a,with MEI of 1.8?4.0a.The SPEI and MEI indiced has a significant resonance period of 1-4a in 1967-1978,indicating that the MEI has significant effects on drought in Hunan Province.Ecological drought time in different ecological regions of Hunan province is obviously different through its frequency analysis.Drought in the Dongting ecological zone has the characteristics of occurrence in spring and autumn,while it would become much weaker in the other seasons.And severity drought occurrence in different ecological zone are respectively spring in Xiang-Gan ecological zone and Wuling ecological zone,and spring and summer in Nanling ecological zone.The spatialization of SPEI at different scales during 2001-2013 show a relatively heavy seasonal drought in the third quarter(July-September)in 2001,2003,2005,2011 and 2013.(2)Through the comprehensive index(EDCI)monitoring,we find that the ecological drought in Hunan Province show significant summer drought characteristics and the close connection between EDCI index and SPEI index indicating that the EDCI composite index is suitable for ecological drought remote sensing monitoring.The drought reached the peak in July and August and alternate between wet and dry in the past 13 years.Mild drought often occurre in April,May and June,and aggravate to moderate in July and August,and then relieve in September.Severe drought occurred in 2003,2011 and 2013,with certain area affected by severe drought.Draught is caused by scarce rainfall in Hengyang and Shaoyang basins and then expanded to other areas,especially in the central region.The year 2003,2011 and 2013 are typical dry years.The proportion of severe drought in Xiang-Gan ecological zone is higher in July-September 2003,which is 81.8%,88.3%and 69.6%,respectively.The moderate ratio of 74%,58.5%and 75.7%in the summer and autumn of 2003 indicated that Xiang-Gan ecological zone is the most severely affected area.From the ratio of drought grading area in various ecological arid regions from July to September 2011,it showed that the proportion of severe drought in Nanling was higher with ratios 51.3%,92.4%and 88.2%than moderate proportions 52.4%,49.1%and 70.7%in Xiang-Gan,respectively.From July to September 2013,it showed that the severe drought in Xiang-Gan accounted for 60.6%in July and in Wuling for 80.2%in August.The correlation analysis between the EDCI index and SPEI index arcross the year,the growing season(April-September)and the quarter(July-September)showed that they correlate with each other.The two mean indices showed that growth season surpass the quarterly and annual index.The monthly correlation analysis showed a high correlation between them from April to September.The mean indices weights are as follows:July>September>August>May>April>June.(3)An optimized CASA model is used to estimate the annual and monthly NPP of Hunan Province from 2001 to 2013,statistical analysis by year,growing season,different vegetation types and ecological zones further verifies the concept of typical drought year in summer and reveal the spatial-temporal variation characteristics of NPP in Hunan Province.In the past 13 years,the total NPP reached a peak of 131.81 Tg C in 2004 and a valley of 103.79 Tg C in 2013.The average NPP in the growing season accounted for 81%of that in the whole year.In the past 13 years,the average annual NPP of natural vegetation is 68.61 Tg C,and that of agricultural vegetation NPP is 26.28 Tg C,but is lower in 2010(25.38),2011(25.33),2003(23.61)and 2013(21.46).The average NPP of natural vegetation accounted for 70%,and the average NPP of agricultural vegetation accounted for 27%.From 2001 to 2013,the cumulative NPP of agricultural vegetation in July and August are affected by drought in 2003,2011 and 2013.In the typical dry year,from July to September,the draught influence on agricultural vegetation varied across different ecological regions.In July,the mean NPP of Dongting is the lowest of 78.72 gc/m2 in 2011,Xiang-Gan of 85.37 gc/m2 in 2013,Nanling of 80.87 gc/m2 in 2013,and Wuling of 90.59 gc/m2 in 2011.(4)The responses of vegetation NPP to EDCI are monitored in terms of time series by year,growing season,quarter and month.The spacial distribution characteristics of four ecological zones of Hunan Province from July to September are compared and analyzed in terms of space,the responsive results of NPP to the two ecological drought monitoring methods are compared and analyzed at month time scale,which show that NPP have better responses to EDCI in most months of the growing season.The annual mean correlation of NPP with EDCI is 0.12,the average growth correlation is 0.25,and the average July-September is 0.57.At monthly scale,the correlation is-0.16 in April,-0.45 in May,-0.14 in June,and 0.52 in July,0.51 in August,and 0.23 in September.Correlation analysis for ecological zone data showed that Dongting and Wuling ecological zones had the highest correlation(0.54)in July,while the lowest correlation were founded in Xiang-Gan zone.In August,Dongting zone has the highest correlation of 0.62.In September,Wuling ecological zone has the highest correlation of 0.28.This change of correlation indicated that spatial variation of ecological drought were changed with seasons in the province.This change may in important in anti-drought action in different regions with different months.The following two points can be identified as the innovation of the study.First,tempo-spatial variations of ecological drought in Hunan Province were systematically examined and revealed from both long time-series meteorological and remote sensing data.Information on the tempo-spatial variation of ecological drought is urgently required to formulate necessary policies and actions for ecological construction in the province.Second,a comprehensive index method was established in the study for monitoring of ecological drought in Hunan province and similar regions in South China.We have validate the method through CASA model.We believe the method can provide alternative for ecological drought monitoring using remote sensing data in the hilly regions with various vegetation covers in South China. |
PDF Full Text Request