Prognosis of water quality using statistical modelling: sulurtank, Coimbatore, India

Pollution of surface water has been growing incessantly in India particularly due to the indiscriminate discharge of waste water from various sources without using any water treatment techniques. Global Water Supply and Sanitation Assessment (GWSSA), 2000 reported that the water related diseases kill a child every eight seconds and are responsible for all illness and death in the world. The scarcity for good quality drinking is increasingly becoming a rising issue and has invited attention. Hence, in this paper an attempt has been made to study about water quality of the Sulur tank which is located at 11.030N and 77.130E of Coimbatore district, India. The tank is heavily polluted due to domestic waste water discharge. The water quality of the tank was analyzed by collecting samples from a point on weekly basis from December 2010 to March 2011. The various physio – chemical and biological parameters such as pH, Electrical Conductivity, Total Dissolved Solids, Total Suspended Solids, Total Hardness, Turbidity, Alkalinity, Nitrate, Calcium Hardness, Magnesium Hardness, Biological Oxygen Demand, Chemical Oxygen Demand, Dissolved Oxygen are analyzed experimentally as per the Indian Standards. The range of the test results obtained are as follows pH (7.2-9.1) , Electrical Conductivity (1330 – 1946.9 µohm/cm) , Turbidity (0.4 – 0.7 NTU) , Total Dissolved Solids ( 770 – 4801.20 mg/l) ,Total Suspended Solids ( 0.3 – 4.5 mg/l) ,Total Hardness ( 48.2 – 237.15 mg/l) , Alkalinity (360 – 900 mg/l) , Nitrate ( 0.56 – 5.17 mg/l) , Calcium Hardness (100 – 156mg/l) , Magnesium Hardness (48 – 105.6 mg/l) , Biological Oxygen Demand ( 18 -29 mg/L) , Chemical Oxygen Demand (98 – 220 mg/l ) , Dissolved Oxygen (17 – 25 mg/l). These parameters are important factors for accessing quality of water. The variation in the test result is due to domestic discharge to the tank and seasonal condition. Using the results, the regression equations are established using SPSS 16. The parameters were correlated using Principal Component Analysis method by which, the highly correlated parameters (r>0.5) were formulated into regression equations. The results useful for the rapid and reliable monitoring measures are essential for keeping a close watch on water quality and healthy environment. The obtained equations are programmed using .NET. Thus the framework helps to predict the future state of water quality easily.