GIS-Based Spatial-temporal Analysis Of Watershed Hydrological Data

Water resources is great significant for regional economic and social development.Grand River is the source of Maderia which is the longest, the largest drainage area and the most complex stream tributary of Amazon. In the Grand watershed, there are 900 million people survive, of which 80% of the water resources for agricultural irrigation. In the basin water resources is shouldering important task such as the agricultural irrigation, the population growth and the economic development.In this paper, spatial and temporal distribution of Grand watershed is studied, which is divided into two parts:First, study on time series; Second, study on the spatial distribution. Take the annual discharge series of three typical stations in Grand watershed as time series' object of study. First of all, the original time series of annual discharge were understood to judge the stability in preparation for later work. Cycles of time series were analyzed by compared Hilbert-Huang Transform and Wavelet analysis to confirm the primary cycle sequence; the trend of multi-year average annual discharge was extracted to determine the trend qualitatively. Twelve kinds of statistical methods were used to identify significant trends quantitatively and the method of Hurst to predict. Spatial interpolation methods were used for special analysis. There are total forty-nine observation stations in Grand watershed. The result was the Kriging method was superior to the traditional interpolation methods through cross-verification.According to the error mean (MEAN), root mean square error (RMS), average standard error (ASE), the standard average (MS) and standard root mean square error (RMSS) value analysis; this paper used ordinary kriging method to interpolation for years of average annual discharge data and contour mapping.Mann-Kendall non-parametric rank correlation test was used to calculate average annual discharge of forty-nine stations, and spatial distribution was mapped.The results showed:The average discharge of 02GA003 Station were non-stationary time series, the upward trend was clearly and will continue in the future; the main period was 7 years,14 years and 36 years. The average discharges of 02GA014 Station was the stationary time series and was no significant trend. The main periods were 9 years and 13 years; the average discharge of 02GA016 Station were the stationary time series and no significant trend. The main periods were 4 years and 8 years. Interannual change was largest at station 02GA016; each month discharge at station 02GA003 was the largest. There were similar for average monthly discharge of three stations that uneven distribution during years with high concentration at period of 3-4 months. The intensive contour showed that discharge at Middle and Lower Grand River changed greatest. The total nine stations were significant trend, there was no significant entirety.