|
基于GIS网络模型发展而来的水利灌溉管理模型(英文原稿)
Development of a model for dam irrigation management based on GIS network models
K.N.Toosi University, Tehran, Iran
Department of GIS, Faculty of Geodesy & Geomathics
Abolghasem Chehreghani, Saadi Mesgari, Mohammad Karimi, Ahmad Seyfi Mastan
Abstract
As a result of population growth and industry development, the consumption of water increases continuously. A large portion of usable water is consumed by agriculture, which requires more efficient ways for both conveyances of water and irrigation. Usually the efficiency of irrigation-water conveyance is too low, meaning that much of the water is either evaporated or absorbed by soil before reaching the user. In this paper, we describe the development of a model for the management of dam-irrigation network. The model is on the basis of network analysis /modeling capabilities of GIS. In this model, the water conveyance efficiency of every segment and consequently every route on the network is calculated. Having the amount of water by farmers on lands, the amount of water needed to be entered to the network is computed. Then, we can compare the whole amount of water needed to be provided with the amount obtainable from the dam discharge. These, in turn, could be used to plan and ration the consumption of water by users and to classify farms according to their priorities for irrigation in an irrigation network.
1. Introduction
The amount of available water resources for irrigation decreases continuously. The reasons for this include population growth, industrial development and global climate changes. According to the available census, agriculture as the main consumer of water uses 92 percent of the 93.1 billion cubic-meter available water (Irani, K.S., 2004).
Agriculture is not only the largest water consumer, but also it uses water with the cheapest price and lowest efficiency (Irani, K.S., 2004). Previous studies in many countries show that the low efficiency of the irrigation systems is mostly the result of weak maintenance and management of irrigation network. This is the main obstacle to the justly consumption of water among farmers. That is, further farms from the water source get much less water than the nearer ones (Irani, K.S., 2004).
The issue of water saving, after a certain point, is that of water saving in irrigation. Thus, irrigated agriculture is increasingly feeling the pressure to demonstrate and improve its performance (Burt, C.M., A.J.Clemmens, T.S.Strelkoff et al., 1997). In this article a method on the basis of the network analysis capabilities of GIS is described to help in a better management of the irrigation network.
2. Modeling of the irrigation network spatial features
Usually there are five types of features in an irrigation network:
·Dam: this is the provider of water to the irrigation system. It can be presented as a point feature having the name, the ID and the amount of water delivery as its attributes.
·Channels: they act as the transporter of water to the farms and are divided into three groups(Khuzestan's energy organization, 2000):
Level-one channel, that receives water from dam and sends it to level-two channels or to other level-one branches;
Level-two channel, starts from a level-one and transfers water to level-three channels;
Channels of level three and four are the branches of the network that deliver the water to individual or groups of farms.
The name and identifier of the channel, its water capacity, the ids of its start and end gates are the attributes that should be linked to channel features.
·Gates/Turnouts: through these features, the water is delivered from higher level channels to the lower level ones. At the end of level-four channels also, the farms receive the water from the farm turnouts. These features are represented as points, having the attributes of ID and the amount of water received (capacity).
·Lands (villages): the main purpose of irrigation network is to transfer water from the sources to these features. When the data about each land is not available, the village could be considered instead. These features are modeled by points. Regarding each cultivation type present at a village, the attributes of cultivation type area and percentage to the whole cultivation of the village are linked to the village.
·Consumption points: the farm turnouts are the consumption points. These points are required to be connected to the villages. The purpose of the irrigation system is to serve these points.
·Other consumers: beside agriculture, usually there are other types of consumers. Name of the feature and its consumption are the main attributes.
·The table of needs for water: this is the only non-spatial part of the model that is connected to consumption points (villages). In this table, the amount of needed water for each cultivation type in area unit is recorded.
3.Description of the model
Irrigation network efficiency is generally defined as amount of water received by consumer divided by the water taken from the source. This includes the three parts of conveyance, field canal and field application efficiency (Khuzestan's energy organization, 2000). Conveyance efficiency is related to the working condition and design of the channels, gates and penstocks. The field canal and field application efficiency depends on the irrigation channels and irrigation system used inside the farms. Therefore, these parts of efficiency are not covered in this article. Here, we calculate conveyance efficiency and call it efficiency.
To define the efficiency of the network, the efficiency of each link should be calculated and considered. To do this, the water delivered to the gate of beginning of the link and the water received at the gate of the end of link are measured. By dividing the output water to the water inputted to the link and multiplying it by 100, the efficiency of the link is extracted.
To calculate the efficiency of the network, on the basis of the shape and connections in the network, it is divided to some regions, each served and fed with a level-two channel. In each region, for every level (levels from two to four) the averages of efficiency of all links are calculated. By multiplying the efficiency averages of all three levels, the efficiency of that region is calculated.
3.1. Calculation of the efficiency The methods of calculating the efficiency of the channels are similar. To get proper formulas, we consider them in two groups of one-part channels and multi-part channels.
·One-part channels
Channels of level four are usually of this type. They consist of one part each. The efficiency, as mentioned, is calculated by dividing the output of the channel by its input water and multiplying it by 100; that is:
E is the water received at the gate of the end of the channel and T is the water delivered to the gate of the beginning of the channel
·Multi-part channel
A channel of this type consists of many parts. In other words, in such a channel there are a few gates. Any such a gate delivers water to lower level channels. Usually channels of level one, two and three are of this type. To calculate the efficiency of such a channel, the channel is divided into simple (one-part) sections. Then for each one of simple sections, the efficiencies are calculated and then multiplied (formulas 2 and 3).
Finally, to calculate the efficiency of the network in a region, the average efficiencies for channels of level two, three and four are calculated separately and then multiplied.
3.2. Defining the water demand for each village
To define the water demand of a village, the table of water demands of plants (cultivation types) is used. By multiplying the water demand per area unit of each cultivation type by the area of that cultivation and addition of the resulted values, the whole water demand of that village is calculated.
3.3. Defining the relation between consumption nodes of the network and farms
To simplify the calculations of water demand, the village as the aggregation of all its farms is connected to the end nodes (consumption nodes) of the network, through a unique identifier. This means that this village is receiving water from that gate/turnout. Having the water demand of villages, the demand of each consumption node can be extracted.
3.4. Calculating the water demand of the whole network
Having the demand of end nodes of the network and adding them in each region, the region water demand can be resulted. By multiplying this with the region efficiency, the real water demand of that region is calculated.
Finally, by adding the real demands of all regions and other non-agriculture consumptions, and multiplying the result by the efficiency of the level-one channel, the real demand of the whole network is achieved. Now this value is compared with the water volume expected to be provided by the water source. If the demand is less than the capacity of the source, then we only need to plan the circling of water (define the irrigation periods) among level-three channels. Otherwise, first, the non-agricultural consumers are cut off; then if the problem remains, the agricultural consumers are classified based on some priorities. Using these priorities, the water portion for each level-two channel obtained and irrigation turns and periods for channels of level three are defined. Although for both these conditions, the portion of water for each channel of level two is defined.
4. Conclusion
In this article, a model is proposed for calculating the efficiency of the irrigation network and using it for the calculation of the water demand. The following results could be mentioned:
·In this method, the conveyance efficiency of the network can be calculated with much more accuracy than the empirical judgements traditionally used.
·The efficiency of each link/channel and region of the network can be calculated separately. Accordingly, proper actions can be taken to improve non-efficient parts.
·The network demand can be calculated more realistically and accurately, which causes better decisions.
·The water dedicated to each consumption type can be calculated.
·In both cases, presence or lack of water shortage, the irrigation cyclic plan and periods for channels of level three should be defined.
|
|