:: Volume 2, Issue 4 (3-2015) ::
jgit 2015, 2(4): 99-119 Back to browse issues page
Modeling Shallow Groundwater Depth Using Hyperion Hyperspectral Imagery
Said Hamzeh *, Abdol Ali Naseri, Seied Kazem Alavipanah, Barat Mojaradi
University of Tehran
Abstract:   (4775 Views)

Current study was conducted in order to finding the best models to estimating groundwater depth using Hyperion hyperspectral satellite imagery in the sugarcane fields located in the southwest of Iran. For this purpose ground water level was measured in 132 observation wells from the beginning of May 2010 till end of September 2010, twice per week, in the Hakim Farabi farming and industrial lands. Moreover, from the other collected information like daily weather information, age and variety of sugarcane, planting and harvesting date of plants, managerial operations such as date and amount of the fertilization, irrigation and drainage information in the Hakim Farabi farming and industrial lands were used. In a same time with measuring the ground data, a hyperspectral satellite image of Hyperion sensor was acquired on September 2, 2010. After applying necessary pre-processing on the image, the changes in the spectral reflectance of the sugarcane under different values of groundwater depths was studied. Afterwards, it was tried to obtain appropriate models for estimating ground water depth. For this purpose, capability of 21 vegetation indices ,related to defferent regions of spectral reflectance of crops, was studied. Besides of these indices three new vegetation indices (SWSI-1, SWSI-2 and SWSI-3) were developed in this study. Results show that, variations of groundwater depths have a significant effect on spectral reflectance of sugarcane. Among the vegetation indices, indices related to water absorption bands or based on a combination of chlorophyll and water absorption bands had the highest correlation with groundwater depth. Obtained models from the two vegetation indices developed in this study (SWSI-1, SWSI-3) and NDWI yield the best results for estimating groundwater depth with R2 of 0.48, 0.48 and 0.47 and root mean square errors of 8.20, 8.25 and 7.98 cm respectively. Conclusions from this study indicate that using hyperspectral satellite imagery to monitoring water table in the sugarcane fields has an acceptable, fast and economical results.

Keywords: Groundwater table, Hyperspectral Imagery, Spectral reflectance, Vegetation Index, Sugarcane
Full-Text [PDF 1205 kb]   (1402 Downloads)    
Type of Study: Research |
Received: 2015/12/6 | Accepted: 2015/12/6 | Published: 2015/12/6

XML   Persian Abstract   Print

Rights and permissions
Creative Commons License This work is licensed under a Creative Commons Attribution-NonCommercial 4.0 International License.
Volume 2, Issue 4 (3-2015) Back to browse issues page