The main objective is to develop decision support system for farmer and irrigation department to gain maximum profit with the maximum use of limited water resources. It also helps to irrigation manager to take decision about the distribution of water among the farmer and city area. It will also help to illustrates implications of the study findings on demands for irrigation and water uses.
Statement of Research Problem
INTRODUCTION TO RESEARCH TOPIC
About Saurashtra (Research Area)
"Saurashtra" is having total land area 6.34 million hectors and agriculture land is 4.17 million hectors. It is said that Water availability plays an important role in agricultural. So the demand of water in irrigation is a crucial task today. To provide water to agriculture area, government has developed more than 120 Dam. About 70-80% of the water of Dam in Saurashtra is currently being used by agriculture. Following Table shows the Crop that is being taken in Saurashtra.
"Crop Taken in Saurashtra using Irrigation Water"
Information of Water Demand
Information regarding water demand in irrigated areas is basic information for the development and implementation of successful water resource management tools given that irrigated agriculture is the largest user of water throughout the Saurashtra. Also, forecasting of water demand is one of the main problems in the design, management and modernization of water supply and distribution systems.
Actually, most pressurized irrigation systems operating on-demand deliver water with the flow rate and pressure required by farm irrigation systems, sprinkling or micro-irrigation, and respecting the time, duration and frequency decided by the farmers. Therefore, they allow farmers to operate their irrigation systems with a large freedom with respect to other types of delivery schedules.
How Existing System Works?
Once a request for water is made it typically takes about 4-7 days to get it at the farm the upstream. Therefore, farmers need to estimate water requirement for the next 4-7 days in advance in order to get it at the farm on time.
However, water requirements calculated for irrigation planning do not always meet the actual use (that is, consumer demand) due to changes in the field environment such as weather conditions and farmers' behavior, which can influence the actual amounts of water used. Actual water management in some irrigation districts is carried out depending only on the experience and knowledge of the administrator although there is always a need to forecast daily water demand.
Drawback of Existing System
A farmer generally either overestimates or underestimates the water requirement. If the requirement is overestimated there will be on farm water loss, whereas if it is underestimated there can be adverse effect on the crop productivity.
Currently there is no reliable tool available to the farmers and irrigation engineer of my study area for estimating future water requirement accurately. Hence, a water demand forecasting technique is crucial for the efficient use of available water. Therefore, having a reliable water demand forecast model can be useful for a farmer to estimate water requirement more accurately. The demand forecasting tool can also be useful for the irrigation managers for estimating water requirement for the whole irrigation area.
Most of the water delivered for irrigation is not always efficiently used for crop production. On an average only 45% of the water is used by crop, 15% is lost during conveyance, 15% is lost in supply channels within the farms and the remaining 25% is lost due to inefficient water management practices [Shentruji Dam Survey Report, May 2011]. Therefore, it is evident that most of the water losses occur at farm level because of inefficient water management practices.
How to Estimate Water Demand?
There are two major approaches for estimating water demand: i) conceptual and ii) system theoretical. A conceptual model predicts the irrigation water requirement based on several factors including soil moisture, seepage, and evapotranspiration. Subsequently irrigation managers use these factors to estimate irrigation water demand for the whole season. However, water requirements estimated at the beginning of the irrigation season may not be the same as the actual water usage due to many reasons such as difference in expected and actual weather conditions and change in farming practices. The second approach for estimating water demand is known as system theoretical approach. In this approach a model is first trained on available data and then used for estimating future water demand. The system theoretical approach is more efficient and accurate than the conceptual approach. Moreover, it can base on easily available data only.
Introduction to Broad area of Research
Data mining is the process of analyzing data from different perspectives and summarizing it into useful information. Technically, data mining is the search for relationship and global patterns that exist among parameters, but are hidden among data.
Data mining, the extraction of hidden predictive information from large databases, is a powerful technology with great potential to help companies focus on the most important information in their data warehouses. Data mining tools predict future trends and behaviors, allowing businesses to make proactive, knowledge-driven decisions. The automated, prospective analyses offered by data mining move beyond the analyses of past events provided by retrospective tools typical of decision support systems. Data mining tools can answer business questions that traditionally were too time consuming to resolve. They scour databases for hidden patterns, finding predictive information that experts may miss because it lies outside their expectations.
The first and simplest analytical step in data mining is to describe the data - summarize its statically attributes (such as means and standard derivations), visually review it using charts and graphs, and look for potentially meaningful links among variables (such as values that often occur together). Collecting, exploring and selecting the right data are critically important.
But data description alone cannot provide action plan. You must build a predictive model based on patterns determined from non-results, and then test the model on result outside the original sample. A good model should never be confused with reality, but it can be a useful guide to understanding your business. The final step is to empirically verify the model. For example, from a database of customers who have already responded to a particular offer? You have built a model predicting which prospects are likeliest to respond to these offers. Can you rely on this prediction?
The data mining is often referenced as K.D.D. Knowledge Discovery in Data Base because in the process of data mining we are mining the data or we are initiated the process of knowledge discovery in data base.
To build an effective model using data mining techniques adequate historical data for the parameters such as crop water usage, crop type and weather conditions are required.
Decision Tree
A decision tree model can be useful for water demand forecasting. Decision tree is a flow-chart-like tree structure, where each internal node is denoted by rectangles, and leaf nodes are denoted by ovals. All internal nodes have two or more child nodes. All internal nodes contain splits, which test the value of an expression of the attributes. Arcs from an internal node to its children are labeled with distinct outcomes of the test. Each leaf node has a class label associated with it.
It recognizes the relationship between the classifying (class) and the classifier (non-class) attributes. A class attribute is an attribute within the data set, which contains the values that are possible outcomes of the record. A decision tree analyses a set of records whose class values are known. In other words, a decision tree explores patterns also known as logic rules from any data set.
ID3 (Iterative Dichotomise 3)
This is a decision tree algorithm introduced in 1986 by Quinlan Ross. The tree is constructed in two phases. The two phases are tree building and pruning. ID3 uses information gain measure to choose the splitting attribute. It only accepts categorical attributes in building a tree model. It does not give accurate result when there is noise. To remove the noise pre-processing technique has to be used.
C4.5
This algorithm is a successor to ID3 developed by Quinlan Ross. In order to handle continuous attributes, C4.5 splits the attribute values into two partitions based on the selected threshold such that all the values above the threshold as one child and the remaining as another child. It also handles missing attribute values. C4.5 uses Gain Ratio as an attribute selection measure to build a decision tree. It removes the biasness of information gain when there are many outcome values of an attribute.
Data Collection and Work Plan
To build the training dataset, I will collect data from three different sources. The first source is the water delivery statements that are obtained from various irrigation division of Saurashtra and provides us with the information about total water usage for a crop growing season by each farm. The second source is the meteorological data that are obtained from the installed weather stations in the study area. The third source is WALMI (WATER AND LAND MANAGEMENT INSTITUTE) at Anand.
The next step after collection of data will be during the data pre-processing step I will compare a few different approaches and finally adapt the most logical one. My training data set will contain attributes on various weather parameters (such as maximum and minimum temperature, wind speed, humidity, rainfall, and solar radiation), soil type, and crop type and water usage. Then I will build a decision tree on the pre-processed training data set in order to extract existing pattern and predict future water demand for the next 7 days. The performance of the decision tree model is compared with a traditional way of estimating water demand using actual evapotranspiration (ETc).
To run the decision tree algorithm, there is a strong need for data pre-processing to prepare good quality data. Data pre-processing takes approximately 80% of the total data mining effort. It is also known that good results can be achieved by using data mining techniques/algorithms only if we have a good quality data. Often the real time data is very inconsistent as it contains many attributes which are not useful for our purpose and have some missing values.
The purpose of data pre-processing is to remove noise from the data, extract and combine the required/relevant attributes from different data sources, make the data reliable and transform the data into our required format. By pre-processing the raw data, it is possible to prepare a good quality data set, which enables efficient and quality knowledge discovery. There are several requirements for data pre-processing tools.
"Data Pre-Processing Tool Requirement"
After data pre-processing next task may include following towards the research work.
Selection of appropriate attributes
Construction of data set (Data Warehouse)
Generating Decision Tree based on the selected Algorithm
Analyze the possible outcomes generated by decision tree
Development of DSS that fulfill demand forecasting based on the data mining technique
Literature Review
For estimating irrigation demand, past studies have mostly adopted fixed water requirement of population, and with one to one mapping to the demand for irrigated land on the basis of population growth.
Instead of that, this study will statistically validates for a curvilinear relationship between irrigation and income level, which suggests that it is not only the level of population, which matters for the demand for irrigated area, but also the level of income and other policies and institutional factors. Hence, projection of the irrigation land (or demand for water uses) should be done taking into account the nonlinear income effects and other substitution processes, which would improve the accuracy and overall performance of the irrigation-forecasting models.
The empirical analysis suggests that the irrigation demand depends upon several underlying factors. For sustainable use of irrigation and water uses for agriculture, these factors need to be considered into the planning and management of irrigation.
Actually, the operational control of most irrigation water distribution systems is based on averaged demand profiles for certain time periods (10-day or weekly periods, normally), which vary according to the development stage of the cropping pattern and the climatic conditions, based on the experience and knowledge of the administrator.
No
Title of Paper/Book
Study review
1
Han, J., & Kamber, M.
Data Mining: Concepts and Techniques,
Basic of Data mining and Techniques
2
Mahmood A. Khan, Md. Zahidul Islam, Mohsin Hafeez
Irrigation Water Demand Forecasting - A Data Pre-Processing and Data Mining Approach based on Spatio-Temporal Data.
How the Proposed System will work. Decision Tree, Data pre-processing, attribute selection, decision support system
3
Inmaculada Pulido-Calvo, Juan Carlos Gutie´rrez-Estrada
Improved irrigation water demand forecasting using a soft-computing hybrid model
How the climate will affect in cropping?
How to make a model to improve demand forecasring?
4
Sudha V, Dr. N.K Ambujam, and Dr K. Venugopal
A Data Mining Approach for Deriving Irrigation Reservoir Operating Rules
Data mining and Decision Rules based on data mining.
5
Godfrey C. Onwubolu
Self-organizing data mining for weather Forecasting
Data Gathering and Data Cleansing
6
M.Kannan, S.Prabhakaran, P.Ramachandran
Rainfall Forecasting Using Data Mining Technique
Data Analysis
7
Petr Aubrecht, Petr Mikˇsovsk´y, and Zdenˇek Kouba
Metadata Driven Data Pre-processing for Data Mining
Data Preprocessing
8
Kampanad Bhaktikul, Rommanee Anujit and Jongdee To-im
Estimation of Crop Coefficient of Corn (Kccorn) under Climate Change Scenarios Using Data Mining Technique
Water Management based on climate change
9
Top 10 algorithms in data mining
C4.5 Algorithm, Decision Tree
Table : Literature Studied
Proposed output
Irrigation managers and farmers can get more accurate future water requirement information.
The experimental results will be providing significant improvement of accuracy in water demand forecasting.
It will compare future water requirement prediction approach with a traditional evapotranspiration based technique.
The data mining based tool developed in the research will crucial for the farmers to make the maximum use of limited water resource.
Upon successful completion of proposed work we will have better utilization of all resources like man power, land, water and fertilizer etc. with better yields in crop production.
The empirical findings of this study contribute to the improved understanding of irrigation requirement, and the search for an answer to the question, how much more irrigation do we really need at any moment of time?
