Estimating evapotranspiration for annual crops in Cyprus using remote sensing
Date Issued
2011
Author(s)
Advisor
Abstract
This thesis provides a novel method for estimating crop evapotranspiration (ETc) on a systematic basis using remote sensing techniques. The major aims of the project are to provide accurate estimations of crop evapotranspiration spatially and to create a method where potential users could monitor irrigation water management through ETc estimation, in the future. Landsat-5 TM and Landsat-7 ETM+ images of the area of interest were transformed into crop evapotranspiration maps, in order to retrieve the value of evapotranspiration for each crop.
To reach this goal, ground truth data, meteorological data, remotely sensed data, modelling techniques and energy balance algorithms were employed and combined. Following the phenological stages of each crop, semi-empirical models were developed, regarding the crop canopy factors of each crop. Vegetation Indices were created from spectroradiometric measurements during crops phenological stages. These indices were used to describe the crop canopy factors, namely Leaf Area Index (LAI) and Crop Height (CH). The developed semi-empirical models were found to have strong correlation coefficients. The models were evaluated with very satisfactory results. The models were finally used to modify the algorithms decided to be used in this project, SEBAL and Penman-Monteith adapted to satellite data.
The direct application of the two crop evapotranspiration algorithms has provided close results but with statistically significant difference. When employing the semi- empirical models, for modifying the two algorithms, the results were even closer and without any significant difference. Maps of crop evapotranspiration were created from the satellite images and the corresponding values were retrieved. The method is applicable for any other satellites with different temporal, spatial and spectral resolution.
To reach this goal, ground truth data, meteorological data, remotely sensed data, modelling techniques and energy balance algorithms were employed and combined. Following the phenological stages of each crop, semi-empirical models were developed, regarding the crop canopy factors of each crop. Vegetation Indices were created from spectroradiometric measurements during crops phenological stages. These indices were used to describe the crop canopy factors, namely Leaf Area Index (LAI) and Crop Height (CH). The developed semi-empirical models were found to have strong correlation coefficients. The models were evaluated with very satisfactory results. The models were finally used to modify the algorithms decided to be used in this project, SEBAL and Penman-Monteith adapted to satellite data.
The direct application of the two crop evapotranspiration algorithms has provided close results but with statistically significant difference. When employing the semi- empirical models, for modifying the two algorithms, the results were even closer and without any significant difference. Maps of crop evapotranspiration were created from the satellite images and the corresponding values were retrieved. The method is applicable for any other satellites with different temporal, spatial and spectral resolution.
File(s)![Thumbnail Image]()
Name
PHD-Παπαδαθίδ Γιώργος .pdf
Size
4.46 MB
Format
Adobe PDF
Checksum (MD5)
11a38ffad9ca47088a3ae9f325644447