| 000 | 09370nas|a2200481 i 4500 | ||
|---|---|---|---|
| 005 | 20260818134542.0 | ||
| 008 | 030627s2002 th uu m rtt 00| a1eng d | ||
| 035 | _a.b11889123 | ||
| 099 | 9 | _aAIT Diss. no. WM-02-01 | |
| 100 | 1 | _aInes, Amor Valeriano M. | |
| 245 | 1 | 0 | _aImproved crop production integrating GIS and genetic algorithms |
| 260 |
_aBangkok : _bAsian Institute of Technology, _c2002 |
||
| 300 | _a207 leaves | ||
| 490 | 1 |
_aDissertation ; _vno. WM-02-01 |
|
| 500 | _aA dissertation submitted in partial fulfillment of the requirements for the degree Doctor of Engineering, School of Engineering and Technology | ||
| 502 | _aThesis (Ph.D.) - Asian Institute of Technology, 2002 | ||
| 520 | _aIn recent years, physically based, field scale agro-hydrological models provided a robust way to understand the cause-effect relationships of the interrelated variables in the soil-plant-atmosphere system. Apparently, this bas increased the flexibility of studying the behaviour of a system under different environmental conditions. However, one main disadvantage of field scale models is their inability to account for the heterogeneity in a system (i.e. a hydrological domain) when the analysis is extended to the regional level. This is important to overcome because of its broad implications to water management studies. In this study, a methodology was developed to address this problem. The field scale Soil-Water-Atmosphere-Plant model (SWAP) was extended to regional application, and then coupled with a Genetic Algorithm (GA), to operate as the core of the developed decision support system, refened here as the WatProdGA model (Water Productivity GA based methodology), which has been used to explore options in water management at the regional level. The integrated model was tested and evaluated in an irrigation system in Kaithal, Haryana, India demonstrating that the methodology provided satisfactory results. The development of the regional model was based on stochastic approach. The physical (e.g. soil, water quality, depth to groundwater) and non-physical properties (e.g. crop and water management practices) of the system were represented by their moments (means and standard deviations), which were used to derive the distributed input data for the regional model. These properties were assumed to be normally distributed in the study area. A parametric bootstrap algorithm was linked with the simulation model to generate these distributed physical and non-physical properties of the system. In this approach, distributed outputs can be generated during simulations by randomly sampling homogenous land units in the system. Obviously, there is the problem on data inputs for the quasi-regional model. Generally, under such a spatial scale, using the classical survey method to derive these distributed properties is clearly impractical. One way to overcome this is to use Remote Sensing (RS) data to extract some information about these properties of the system. A part of the methodology is to develop a system characterization method by means of a regional inverse modeling. GA was integrated with the regional model to search for a solution during the inverse modeling. The spatial and temporal distributions of evapotranspiration (ET) were considered as fitting criteria in the search. The estimated regional ET from RS data were used as measured values of ET over the area, and the resulting ET distributions from the simulations (using the generated parameters from GA as input data), were matched with the measured ET values in the process. GA fitted the generated ET with the measured ones by breeding the appropriate parameters (moments) for the study area. Two Landsat 7 ETM+ data taken on February 4 and March 8, 2001 were used in the analysis. The spatial distributions of ET were calculated from a GIS based remote sensing model called SEBAL - Surface Energy Balance Algorithm for Land. The theory behind the use of ET as fitting criterion is based on the idea that the crop and water management practices, as well as the soil hydraulic properties, could influence the spatial and temporal variability of ET over a region. Consequently, these variations could provide some information on the properties of a system. To verify this idea, a field scale inverse modeling was conducted using lysimeter data. The field scale SW AP model was integrated with GA to estimate selected soil hydraulic parameters (i.e. a, n, Ksai. and 8sat of the Mualem-Van Genuchten equations) where ET and soil water were evaluated as search criteria. Results showed that ET was found to be weak in predicting the soil hydraulic parameters compared to soil water. Obviously, because, soil water has a direct relationship with these soil hydraulic parameters. Notwithstanding, the predicted parameter values using ET could produce a good fit of the soil water profiles compared to the measured values in the numerical case (forward-backward simulation). In the experimental case (where actual lysimeter data were used), soil water was still more superior to ET. The results on ET, on the other hand, are worst than in the numerical case. Further analysis showed that this discrepancy could be attributed to the uncertainty in ET estimation itself, data and model errors. Proper definition of the sensitive parameters in ET calculation could have improved these solutions. Despite the weakness of ET in estimating the soil hydraulic parameters, its application to the regional scale is attractive because it can be estimated with reasonable accuracy using RS data e.g. from SEBAL analysis. The regional inverse modeling was applied to characterize the properties of the investigated system. The experience gained from the field scale inverse modeling had served as a guide to the system characterization. This resulted to some reasonable outputs based on the comparison of the predicted parameters and the observed data in the irrigation system. The GA based decision support system called WatProdGA, is composed of four main modules (1) the regional SWAP model, (2) data module, (3) Genetic Algorithm and (4) the output module. The link between the regional model and GA is data, which can be provided by the system characterization algorithm, which is the reverse mode of the Extended SW AP-GA linkage. This can be done as a stand-alone procedure. The decision support system uses the forward mode of the simulation-optimization linkage to explore options in water management. The decision variables in the water management model were the moments (means and standard deviations) of the crop and water management practices, represented here by the sowing dates and an irrigation scheduling criterion. The goal was to find the optimum management strategies that would maximize the productivity of wheat in the system under different environmental conditions. In the implementation, five scenarios were explored: when the average water supplies available in the system are 200, 300, 400, 500 and 600 mm, which cover a range of situations when water supply is scarce and when it is abundant. The 300-mm scenario represents the present status of the system. Results showed that during times of water scarcity, equitable water distribution could increase the overall performance of the system. The limited water supply has to be spread equally to the system constituents. To achieve this, a wide distribution of sowing dates is necessary and the sowing dates should be earlier. Likewise, during periods of water abundance, the available water has to be spread equally to the constituents to achieve an optimum productivity in the system. The farmers, however, have higher degrees of freedom in their planting activities. The regional model and decision support system developed in this study can contribute significantly to the spatial and temporal analyses in regional water management studies. | ||
| 650 | 1 | 0 | _aGeographic information systems |
| 650 | 1 | 0 | _aGenetic algorithms |
| 650 | 1 | 0 | _aCrop improvement |
| 700 | 1 |
_aGupta, Ashim Das, _eChairperson |
|
| 700 | 1 |
_aLoof, Rainer, _eCo-Chairperson |
|
| 700 | 1 |
_aDroogers, Peter, _eExamination committee |
|
| 700 | 1 |
_aClemente, Roberto S., _eExamination Committee |
|
| 700 | 1 |
_aHonda, Kiyoshi, _eExamination Committee |
|
| 700 | 1 |
_aSingh, Vijay P., _eExamination committee |
|
| 710 | 2 |
_aJapanese Government Scholarship, _eScholarship donor |
|
| 810 | 2 |
_aAsian Institute of Technology. _tDissertation ; _vno. WM-02-01 |
|
| 856 |
_3Full-Text _uhttp://203.159.5.9/ait-thesis/detail.php?q=B11743 |
||
| 907 |
_a.b11889123 _bmnait _cu |
||
| 902 | _a240513 | ||
| 998 |
_b0 _c030627 _dm _ea _fu _g0 |
||
| 945 | _lmnait | ||
| 945 | _lmnait | ||
| 945 | _lmnarc | ||
| 942 | _c22 | ||
| 942 | _c40 | ||
| 942 | _c20 | ||
| 909 |
_aBarcode : 30050120554901 _bCREATED : 2012-08-24 _cRECORD # : i12690934 _dLPATRON : 1025965 _eLCHKIN : 2016-05-28 _f# RENEWALS : 0 _g# OVERDUE : 0 _hIUSE3 : 0 _iTOT CHKOUT : 1 _jTOT RENEW : 1 |
||
| 909 |
_aBarcode : 30050120234215 _bCREATED : 2016-04-04 _cRECORD # : i12921713 _dLPATRON : 0 _eLCHKIN : - _f# RENEWALS : 0 _g# OVERDUE : 0 _hIUSE3 : 0 _iTOT CHKOUT : 0 _jTOT RENEW : 0 |
||
| 909 |
_aBarcode : 30050121017551 _bCREATED : 2020-06-07 _cRECORD # : i13280181 _dLPATRON : 0 _eLCHKIN : - _f# RENEWALS : 0 _g# OVERDUE : 0 _hIUSE3 : 0 _iTOT CHKOUT : 0 _jTOT RENEW : 0 |
||
| 999 |
_c58632 _d58632 |
||