基于不同旱情指数的石羊河流域春旱监测研究

Spring drought monitoring in Shiyang River Basin based on different drought index

关键词:  环境星  石羊河流域  旱情监测  土壤湿度

Key words:  environment satellite  Shiyang River Basin  drought monitor  soil moisture

摘要:

【目的】 对石羊河流域的旱情进行监测，找出更能体现区域旱情变化趋势和适宜石羊河流域的监测指数，为业务化干旱监测打好基础，提供科学手段，也为该区农业生产和防灾减灾提供决策依据。【方法】 文章基于HJ-1A/B的CCD和IRS数据，针对2018年石羊河流域出现的一次较为严重的春旱，采用3种遥感旱情监测指数（TVDI、VSWI、VHI）进行监测，并结合实测土壤湿度进行相关性分析并建立模型。【结果】 3种指数均能反映出研究区域旱情的发生发展及变化过程。其中，3—4月VSWI指数与土壤湿度的相关性最高，相关系数分别为-0.85（P<0.001）和-0.89（P<0.001）；5月旱情缓解后，TVDI指数与土壤湿度的相关性高于VSWI和VHI指数，相关系数为-0.76（P<0.001）。将监测点划分为水浇地和旱地两种类型，TVDI指数适用于反演水浇地的土壤墒情；对于旱地类型，3种指数均可以用来反映旱地土壤墒情。建立土壤湿度与3种遥感指数的关系模型，3—4月VSWI指数建立的模型调整后的R²最高，分别为0.72和0.79，模型的平均绝对误差和均方根误差均低于10%。5月旱情得到缓解后，TVDI指数建立的模型调整后的R²在0.58以上，高于VSWI和VHI指数。【结论】 TVID指数更适用于墒情较好的区域进行土壤湿度的反演，VSWI指数则适合在干旱时期进行干旱监测和反演土壤湿度。

Abstract:

Purpose In order to find the best and most suitable index to reflect the changes of drought in Shiyang River ，three indexes were used to monitor the drought based on the vegetation index（NDVI）and land surface temperature（Ts）. It can be used for setting a foundation for the future operation drought monitoring and providing scientific methods for getting accurate drought distribution. Based on this，providing decision-making basis for agricultural production and prevention and mitigation of disaster in the area.Method The data of environment satellite HJ-1A/B CCD and IRS was used to analyze a severe spring drought in 2018 in Shiyang River basin based on three indexes. Correlation between measured soil moisture and indexes were analyzed and then built models.Result Three indexes can reflect the development and changes of the drought in the study area. The vegetation supply water index（VSWI）had the highest correlation with soil moisture in March and April and the correlation coefficient were -0.85（P<0.001）and -0.89（P<0.001）respectively.After the drought eased in May，the correlation between the temperature vegetation drought index（TVDI）and soil moisture was higher than VSWI and vegetation healthy index（VHI）and the correlation coefficient was-0.76（P<0.001）. Dividing the study area into irrigated land and dry land，TVDI was more suited to reflect the soil moisture in irrigated land and in the dry land，the three indexes were all suitable. Soil moisture inversion models were built based on three indexes，the adjusted R² of VSWI in March and April were the highest，0.72 and 0.79 respectively. The mean absolute errors and root mean square errors of the models were less than 10%. After the drought eased，the adjusted R² of TVDI was higher than VSWI’s and VHI’s.Conclusion These results indicated that TVDI was more suited to inverse the soil moisture in better soil moisture conditions and VSWI was more suited in the drought.