| In the summer of2006, Sichuan-Chongqing region suffered the worst drought in recent50years, combined with the persistent drought occurring in Southwest of China from2009to2010, both of which caused great losses in crop yield. Paddy rice is the staple in this region. Quantitative assessing and monitoring agriculture meteorological drought has great significance on agriculture structural adjustment and policy-making with respect to disaster prevention. Recently, few studies have focused on meteorological drought risk assessment in Sichuan-Chongqing region. This paper uses the50-year meteorological data obtained from43meteorological stations, and the information associated with rice yield, rice growth area and disaster statistics to explore agriculture meteorological disaster risk assessment in the Sichuan-Chongqing region with quantitative analysis of rice yield losses in2006.This paper selected3first-grade indices, including the hazard risk (H), the order vulnerability (V) and the disaster resistance (RE) and12second-grade indices to establish the comprehensive assessment index (R). In the aspect of remote sensing monitoring, TRMM3B43(Tropical Rainfall Measure Mission) was used to calculate monthly precipitation anomaly and cumulative monthly precipitation anomaly in order to indicate the spatial distribution of meteorological drought from2000to2012in studied region. The19typical drought periods were selected according to TRMM data. Three temperature-vegetation drought indices (TVDIN, TVDIE, and TVDIM) were constructed to analyze their feature spaces, and their relationship with precipitation trends and soil moisture data in10cm and20cm depth from98meteorological sites. Spatial comparison between TVDIM-based soil moisture and TRMM-based precipitation data from these selected19typical drought periods was made. Given the relationship as "precipitation-soil moisture-vegetation growth" the lag-response of anomaly vegetation index (ANDVI) to precipitation and soil moisture was analyzed. Three methods as Lagrange interpolation method, Linear moving average method and Average yield-reduction method (with rice area extraction, yield assessment, and drought-suffering area estimation) were used to estimate rice yield losses in2006.This paper focuses on agriculture meteorological drought risk assessment, remote sensing monitoring and rice yield losses assessment, all of which are crucial in agriculture drought study area. The key findings are as follows: (1) The highest R values were found in Chengdu, Deyang, Chongqing and Suining city, et al. Various factors could lead to the high R values, and the high order vulnerability was the main contributor. The areas in the western and northern region of Sichuan, such as Aba and Ganzi Tibetan Autonomous Prefecture have the lowest R values. Rice yield loss model was used to validate the agriculture meteorological drought risk model used in this study. These two models had a significant correlation (R2=0.45, P<0.05).(2) The data derived from TRMM and observed precipitation amount were significant correlated (P<0.001). TRMM-based precipitation anomalies showed that the severe drought occurring in June to August of2006were mainly found in areas such as Yibin, Shapingba, and Suining city. The high temperatures and shortened precipitation were key reasons. The drought covered all the areas except for the northern part of Sichuan. This result is well corresponded to those obtained in Chapter one. TRMM-based precipitation anomalies can well reflect spatial distribution and temporal evolution of drought in Sichuan-Chongqing region, especially for2006and2009-2010. The spatial distribution of TVDIE-based soil moisture in most observed month was matched or similar to those of TRMM-based precipitation anomaly. Then the spatial distribution of anomaly vegetation index (ANDVI) in studied region was established. The correlation coefficients of ANDVI and TRMM-based precipitation anomaly reached to0.32and0.33in40th and48th day, respectively. Meanwhile, the correlation coefficients of ANDVI and TVDIE-based soil moisture reached to0.35in16th day, which clearly shows a lag-response among them.(3) The expectation yield curve obtained by the Lagrange interpolation method was over the actual yield curve. The rice yield losses in Sichuan and Chongqing region by this method are2.73and1.39million tons, respectively (with a total of4.01million tons). The average (1949-2011) rice yield losses amount and percent in Sichuan were1.18million tons and8.35%, respectively, while these data were0.423million tons and7.6%in Chongqing from1997to2011. The rice yield loss amount (million tons) and percent(%) in2006was much higher. Trend yield curve obtained by Linear moving average method fluctuated among the actual yield curve. The estimated rice yield losses in Sichuan by this method were1.56million tons. It shows that the expectation yield obtained by Lagrange interpolation method may lead to over-estimation due to fact that the selected expectation yield is too ideal, while Linear moving average method may under-estimate the results since the meteorological yield cannot be well separated from the actual yield. The extracted rice areas in Sichuan-Chongqing region by remote sensing technology were3.5×106ha with the relative errors about15%when compared with data from National Bureau of Statistics. The no harvest areas, drought-occurred areas and drought-induced areas were estimated to be8.1×103ha,45.2×103ha and2.67×106ha, respectively in studied region based on vegetation anomalies and results obtained above. The rice yield losses according to the Average yield-reduction method are3.02million tons. |
PDF Full Text Request