Predicting hydrological response to forest changes by simple statistical models: the selection of the best indicator of forest changes with a hydrological perspective

International Conference on Energy Engineering and Environmental Protection 2016

能源学;生态环境科学

Ning, D.^1 ; Zhang, M.^1,2 ; Ren, S.^1 ; Hou, Y.^1 ; Yu, L.^1 ; Meng, Z.^1

School of Resources and Environment, University of Electronic Science and Technology of China, Chengdu

611731, China^1

Research Institute of Forest Ecology, Environment and Protection, Chinese Academy of Forestry, Beijing

100091, China^2

关键词: Correlation analysis;    Ecological protection;    Hydrological impacts;    Hydrological prediction;    Hydrological response;    Independent variables;    Multiple linear regression models;    Waterresource management;   

Others  :  https://iopscience.iop.org/article/10.1088/1742-6596/52/1/012059/pdf
DOI  :  10.1088/1742-6596/52/1/012059

学科分类：环境科学（综合）

来源: IOP

【 摘 要 】

Forest plays an important role in hydrological cycle, and forest changes will inevitably affect runoff across multiple spatial scales. The selection of a suitable indicator for forest changes is essential for predicting forest-related hydrological response. This study used the Meijiang River, one of the headwaters of the Poyang Lake as an example to identify the best indicator of forest changes for predicting forest change-induced hydrological responses. Correlation analysis was conducted first to detect the relationships between monthly runoff and its predictive variables including antecedent monthly precipitation and indicators for forest changes (forest coverage, vegetation indices including EVI, NDVI, and NDWI), and by use of the identified predictive variables that were most correlated with monthly runoff, multiple linear regression models were then developed. The model with best performance identified in this study included two independent variables -antecedent monthly precipitation and NDWI. It indicates that NDWI is the best indicator of forest change in hydrological prediction while forest coverage, the most commonly used indicator of forest change is insignificantly related to monthly runoff. This highlights the use of vegetation index such as NDWI to indicate forest changes in hydrological studies. This study will provide us with an efficient way to quantify the hydrological impact of large-scale forest changes in the Meijiang River watershed, which is crucial for downstream water resource management and ecological protection in the Poyang Lake basin.