Authors
Abstract
1- Introduction
The drought is considered to be the most environmentally harmful phenomenon among the set of destructive meteorological phenomena and it has devastating effects on economic, social, political, agricultural and ecological issues. To this end, predicting drought is one of the proper ways to deal with this natural phenomenon and reducing its devastating effects. Also, since a large percentage of the protein and energy requirements of Iran's population are provided from strategic crops such as wheat and barley, water scarcity and droughts have put food security into trouble to the extent that if there is no planning, food security will face many challenges. The sensitivity of the agricultural sector in the event of a drought in line with the above-mentioned gap, so that if a short-term drought coincides with critical stages of plant growth, will have detrimental effects on the production rate all of which indicate that a particular attention should be given to conducting applied research in the field of agricultural drought monitoring and prediction. In this regard, it is required that the degree of sensitivity to drought areas, planting and cultivation of some crops or changes in cropping patterns and crop planning in different areas be detected according to monitoring associated with drought and then the needed zoning be done. Production performance simulation and elements of the water balance regarding the spatial variability of soil, meteorological parameter, crop cultivation calendar, irrigation water resource, plant physiological components all can be considered effective in evaluating drought at exact scale.
Research on agricultural drought monitoring and evaluation of production dates back to 1930 and the majority of studies have been carried out in the US. These studies revealed that in order to evaluate drought, a definite index should be defined according which the presence or absence of drought and the degree of severity be evaluated. Simulating the plant growth which is exposed to drought and also simulation of the water balance of a region are other techniques to estimate drought. Combining meteorological parameters and soil, water and herbal element's data together utilizing water, soil and crop growth simulation models allow the possibility to apply it as an effective tool to predict drought and production. In the past few decades, the physical Agro hydrological models such as WOFOST and SWAP and empirical models like VSM have been developed all of which truly simulate the processes of plant growth and soil water balance. In this regard, combining these models with remote sensing techniques for simulating regional crop yield and water balance components have been the attention of many researchers. According to the results obtained from various regions, it seems that utilizing SWAP as a physical Agro hydrological model is a good technique to estimate the prediction of drought in the region under study to the extent that by applying this technique the user will be able to predict wheat production performance at different intervals before harvest. The present research aims to develop an assessment system utilizing the SWAP model in order to predict drought according to wheat production performance index in dry and irrigated land under cultivation on Isfahan with an area of approximately 340,850 acres. And finally by networking the data collection, the leaf area index (LAI) of dry weight of the product is networking-like produced.
2- Methodology
Drought is conceptually defined as a consecutive series of rainfall lack along with improper temperature for plant biology lead to reduced production performance. In the operational definition of drought also, the water balance of the plant associated with daily precipitation and evapotranspiration are compared and soil moisture depletion speed and reduction of plant water is determined.
Regarding the importance of strategic crops such as wheat, agricultural drought prediction is one of the best performance indices of this product in areas under rain-fed and irrigated since such an index predicts drought effects on yield production months before harvest time through combining physical Agro hydrological model as SWAP and meteorological and satellite data. Input data required by this system from satellite images include predicted short- and long-term monitoring data, max and min average temperature, wind speed at 2 meters, actual vapor pressure of water, radiation and rain on a daily basis in a year, soil data, land use maps and elevation and vegetation indices. The software employed includes MATLAB and Arc INFO programming. Model for agricultural drought prediction assessment system is based upon the way in which the area under study is separated into a spatial network size from 500 to 500 meters then, the model is performed for each network unit and for each crop year. Therefore, the implementation timeframe of the model divides into two intervals: the first interval (1) includes crop planting date until run-time of the model and the second interval (2) begins from the run-time of the model by the user until the harvest time which is considered as the prediction interval of the model. This system comprises of four sections:
1. Automatically downloading and updating the data
2. Processing the data
3. Extraction of computational units
4. Modeling and simulation of plant growth and crop production
Input data of the system, each using a certain method for a given network size of (500-500) are regionalized and by combining regionalized informational layers together, the computational unit in the system is derived. After processing the data and preparing the data set for each computational unit, the SWAP Agro hydrological model is coupled with computational units in order to simulate the wheat production performance.
By networking this data set, the output of the system was provided in terms of spatial maps of predicted dry weight production performance of the grain wheat and the results obtained were compared to actual performance of the wheat yields across eight towns in Isfahan.
3- Discussion
The system was implemented as an example for the crop year of 2010-2011 on May, 9th and the obtained results were fully presented in terms of spatial maps of rainfall, temperature and radiation meteorological parameters, normalized difference vegetation parameters (NDVI) and leaf area index (LAI), components of water balance in the soil and also a spatial distribution map of wheat yield performance prediction. While applying the system, the spatial distribution map of wheat yield performance prediction was prepared in two forms of total dry weight and dry weight of wheat grain. Figure (1) demonstrates the spatial distribution of dry weight yield performance prediction of the total wheat and the grain weight Kg per ha in the area on 10 May 2011 equal to 20th of Ordibehesht, 1390. Moreover, in figure (2), relative frequency of total dry matter weight production and dry matter weight of wheat grain in the computational units of the area on 10 May 2011 is considered. According to figure (1), dry matter weight production of the total wheat in about 39% of the pixels within is 12000 ha on average while the dry weight production of the wheat grain in about 38% of the pixels within is 4000 ha on average.
Fig (1): the spatial distribution of dry weight yield performance prediction of the total wheat and the grain weight Kg per on 10 May 2011
Fig (2): Histogram of the total dry matter weight and wheat grain weight yield performance prediction per kg
4- Conclusion
In order for assessment and verification of the final output of agricultural drought prediction system including the spatial distribution of dry weight yield performance prediction of the wheat grain, this output was compared with actual performance values of wheat production shown in table (1-A)across 8 towns in Isfahan province in the crop year 2010-2011 and comparative results were then presented in table (1-B). According to the data obtained, the difference lies between the actual as opposed to anticipated wheat production performance values ranges from 78.1 to 88.2%.
Table (1): Comparison of the average actual and predictedperformanceby the system in Isfahan Province (Year: 2011)
Table (1-A)
towns
city Komayni Chadgan
Tiran and Koroon city Shahin and Meyme Barkar Ardestan Aran and Bidgol
Esfahan
prediction
wheat wheat wheat wheat wheat wheat wheat wheat
Performance in surface kg per hectare 5600 4000 4200 4800
5200 4700 4800 5600
Table (1-B)
towns city Komayni Chadgan Tiran and Koroon city Shahin and Meyme Barkar Ardestan Aran and Bidgol Esfahan
predicted performance kg per ectare
4758 3386 3297 4010 3313 4039
3971 4374
Difference predicted and actual values% 85 847 785 835
81 859 883 781
Regarding the results obtained from the SWAP physical Agero hydrological model in lands under cultivation in Isfahan, it was revealed that employing this technique in order to assess the prediction of agricultural drought in the area under study is appropriate and practical and the user will be able to anticipate the wheat production performance in various intervals before the harvest employing this technique.
Keywords
)
