1Research Scholar, Department of Water Resources Development and Management, IIT Roorkee, Roorkee, Uttarakhand, India
2Scientist F, National Institute of Hydrology, Roorkee-247667, Uttarakhand, India
3Professor, Department of Water Resources Development and Management, IIT Roorkee, Roorkee-247 667, Uttarakhand, India
*Corresponding author e-mail: schakraborty714@gmail.com
Online published on 11 September, 2014.
In this paper, an attempt has been made to study the spatial and temporal variability of rainfall at Seonath sub basin in Chhattisgarh State (India) for 49 years (1960–2008). To detect correlation Lag-1 autocorrelation coefficient was used. Mann-Kendall (MK) or Modified Mann-Kendall (MMK) (non-parametric) and Spearman's rho test (parametric) were applied to detect the trend. Sen's slope was used to detect trend magnitude. The CUSUM and cumulative deviations test were applied to detect change points. ArcGIS 9.3 environment is used to investigate the spatial trends. The Coefficient of Variation (CV) was used for variability analysis. According to both the tests decreasing trend was found in annual and seasonal rainfall series for the whole river basin. The maximum decreased annual rainfall was found for Bodala station (−13.8mm/year) and minimum at Simga station (0.1mm/year). The results indicated an overall downward trend in annual and seasonal rainfalls by both the trend methods. The most probable year of change is 1980. The trend test for two separate periods before (1960–1980) and after (1981–2008) the change point indicated, increase in rainfall before the change year whereas decreasing trend after the change point. The CV revealed the inter-annual variability was high for the entire basin (22.03%).
Rainfall, Trend analysis, Mann Kendall test, CUSUM test, Cumulative deviations