Data Mining through Modelling in Irrigation Commands

Introduction

Supplying water naturally or artificially to soil, essentially for the growth of plants of crops is the basic purpose of the activity called “Irrigation”. Using nature, it could be carried out through flow irrigation using perennial and/or flood irrigation system. Uncertain potentiality of the erratic natural systems led to the investments in man-made resource mobilisation aimed at food safety. Artificial ways of doing it comprised of construction of storages and canal conveyance systems sometimes including lifting mechanisms. Generations survived with such a use of natural resources like land, water and air, however synthesised as exploitation once production started losing its grip and number of thrust habitants started increasing. To spare more water for drinking, it became necessary to maximise the productivity of the irrigated water through unlimited use of fertilisers and pesticides. Sprinkler irrigation made the major breakthrough due to its consumptive water spreads and almost no draining scope. Canal Systems were converted in to water tight lined conveyance instruments. Possibility of vertical flows now exist only when precipitation exceed the initial soil moisture requirements and has the capacity to infiltrate the soil masses having voids filled with excessive fertilisers and pesticides.

Soil consists of finely divided organic matter and minerals formed by disintegration of rocks as well as soil colloids and soil solutions. Sandy soil is known as light soil, loam as medium or normal soil and clay as heavy soil. Different types of soils are good for different crops. Crops requiring more water are nicely grown in heavy soils like clay. Light soils containing sand through which water quickly passes down can grow crops with lesser water requirements. Normal soils are well drained and ventilated, ideal for crops with average water requirements. The salts commonly met with in the soil crust are the sulphates, chlorides, carbonates and nitrates of sodium, potassium and magnesium and also chlorides and nitrates of calcium. If the percentage of the total harmful salts does not exceed 0.18%, the yield from the soil is not generally affected. But if it exceeds 0.25%, the soil becomes infertile. For normal yield of crop, the PH value of the soil should lie between 6 and 8.5. When it exceeds 11, the soil becomes useless and unable to grow any crop. Interesting to note that the fall out of the irrigation efforts, if in one hand could ensure the food security, on the other hand agricultural commands gradually became infertile in many places visibly diagnosed as water-logging and salinity of soil causing efflorescence. Soil erosion due to floods, rains and winds segregates the chemical fertilisers with top soil, not allowing it to be consumptively used and improve fertility. The United Nations Food and Agriculture Organization (FAO) estimate that 60–80 million hectares are affected to varying degrees by water-logging and salinity (FAO, 1996). It must have been increased many folds since then, despite many research and development plans to mitigate the effects. Dry land salinity is an added complexion of the problem, where water supply has limited role to play. However different estimates have shown that worldwide only 20 to 24% of the irrigated lands suffer infertility due to salinity and water-logging (Khan, Rana, & Hanjra, 2007), also reports magical recovery in the situation in India since 1991.

Irrigation command areas do exhibit such condition invariably and many times nobody knows whom to blame. The process of soil reclamation is costly and difficult in many occasions. Leaching process requires open or covered drains connected to some natural drains or river, tracing of which is a hard job in present groundwater table conditions. Scrapping of the top surface even for a minimum depth of 1 cm to 2 cm requires waste lands to throw out the useless soil. Addition of chemicals like gypsum works only in the cases of sodium carbonate etc., otherwise bioremediation is the present day technique, widely used; still not acceptable in general. Mathematical implications of command area problems concentrate more towards the water and salt balances in a domain and less to deal with the flow and transport phenomena. Suggesting artificial recharge or conjunctive use of water resources as a permanent solution many times aggravated the problem rather than providing rescue. Earlier lack of data used to be in the core of not able to diagnose the problem. Now that enough hydrological data base have been generated and accessible to handsome amount of users, need is to realise where the possible use and threat of data mining occurs and how the same could be delimited to have appropriate use of conceptual modelling in the irrigation command areas.