Fusion of spatio-temporal remotely sensed evapotranspiration by data assimilation for irrigation performance | |
| Author | Chemin, Yann H. |
| Call Number | AIT Diss. no.RS-06-5 |
| Subject(s) | Irrigation--Remote sensing Evapotranspiration--Remote sensing Remote-sensing images |
| Note | A dissertation submitted for the partial fulfillment of the requirements for the degree of Doctor of Technical Sciences |
| Publisher | Asian Institute of Technology |
| Abstract | Monitoring of water consumption in irrigation systems has become increasingly important for water managers in the actual trend of integrated water management. Low spatial resolution (LSR) satellite remote sensing has already proven the capacity of monitoring evapotranspiration over large areas at high-temporal frequencies, by which monitoring for large irrigation systems can be satisfied. However, smaller pixel size is still required for more local management, while keeping return period within few days. High spatial resolution (HSR) satellite imagery is indeed available for calculation of evapotranspiration, and has been used in many studies already. However, its practical return period is a major drawback to its implementation for monitoring irrigation systems. This research is perusing into the use of genetic algorithms to assimilate parameters of an agro-hydrological model called SWAP for each of the pixels of HSR images contained into one single pixel of a LSR multi-temporal image. The methodology developed and experimented here is trying to take advantage of the spatial content of HSR images and the temporal content of LSR images by fusing them by the process of data assimilation. This methodology differentiates itself from regular fusion models by the major fact that it is `not a spectral fusion but a' simulation-based fusion using data assimilation as a conductor. The total errors of the crop model (SWAP) to both of high and low resolution RS ETa was evaluated in a single objective function which was minimized by a newly developed GA based optimization algorithm. The optimization algorithm was named "double layer GA optimization" which essentially is a set of two GA, one being implemented at every generation of the other like a recursive relationship. Daily and high spatial resolution (180m x 180m) actual evapotranspiration data was generated using the assimilated crop model in a low resolution RS pixel (1km x lkm). The data assimilation experiment with corrected LSR data meets the accuracy required for HSR average differences of 0.05 and -0.07 cm/day. The accuracy is less at LSR with an average difference of 0.15cm/day, which is not satisfactory. The spatial content of HSR images and the temporal content of LSR images can be assimilated. The main advantages found are that satellite images from any date can be used for both HSR & LSR, even if it would be preferable to have at least two HSR images per cropping season. Additionally, that this methodology can be used to potentially generate, for each HSR locations, any time series data that come as output of the simulation model. The output is not limited to ETa but crop growth, yield, water stress, or whatever model output of the simulation. Thus this framework can be used for monitoring / predicting / scenario simulation / optimization in various aspects. The main limitation found at HSR is that the model parameterization does not describe enough the HSR variations of ETa, and may need refinement in variable and fixed parameters |
| Year | 2006 |
| Type | Dissertation |
| School | School of Engineering and Technology (SET) |
| Department | Department of Information and Communications Technologies (DICT) |
| Academic Program/FoS | Remote Sensing (RS) |
| Chairperson(s) | Honda, Kiyoshi |
| Examination Committee(s) | Tripathi, Nitin Kumar;Kumar, Sivanappan;Surat Lertlum;Oki, Taikan |
| Scholarship Donor(s) | Self-supported |
| Degree | Thesis (Ph.D.) - Asian Institute of Technology, 2006 |