ASSESSMENT OF BASEFLOW FOR MICRO WATERSHEDS IN THE MIDLANDS OF KERALA

Abstract

A study was conducted to quantify baseflow deriving from the micro watersheds of river Bharathapuzha lying near Valanchery town in the Malappuram district of Kerala, named as Perassannur watershed (W1) and Painkannur watershed (W2). Various baseflow separation methods such as graphical methods including straight line, fixed base and constant base methods; UKIH and digital filter methods including Lyne and Hollick (LH) and Chapman and Maxwell (CM) methods were used for the quantification of baseflow. Their performance evaluations were done to find the best suitable method for the selected study area. Micro watersheds were delineated from SRTM DEM using GIS software and their geographical areas were found for both the watersheds. Daily rainfall and stream flow were measured for both the micro watersheds by digital rainfall event recorder and area velocity method respectively. Morphometry of the study watersheds in terms of areal, linear and relief aspects; and land use and soil analysis were made in the GIS platform to gain insights into the parameters influencing the baseflow generation process. Graphical and statistical comparisons were performed among the methods to identify the best filter parameters for different filtering methods to make the baseflow separation process more accurate and acceptable. Morphometric analysis showed that the maximum stream order of Perassannur is 5 and Painkannur 4, and both of them have dendritic drainage pattern. Most of the area of the watersheds was covered by cropland and paddy fields, which facilitated the baseflow production. The value of N in the UKIH method was found as 4 in Perassannur and 3 in the Painkannur watersheds. Filter parameters of the digital filter method were determined using the procedure given by Cheng 2016 and estimated as 0.70 for Perassannur and 0.79 for Painkannur, and compared their performance with the often used filter parameters with the values of 0.850 and 0.925 in which 0.925 gave better results than the computed values for both watersheds. Estimated baseflow values for each method were statistically compared with the observed base flow values to find the best suitable method. Among the different filter methods, the LH method showed better accuracy than other methods while the variable slope method has come out as the most inaccurate procedure to estimate baseflow in the study region.