Rice crop mapping in the Mekong River Delta using TerraSAR-X radar remote sensing data
Vietnam is one of the biggest rice exporting country in the world in which the Mekong Delta accounts for over half of the country's rice production. Rice is the dominant plant grown in Vietnam and Asia. Therefore, it becomes necessary to monitor and estimate rice acreage in a large area for the management and planning of the managers.
The advantage of radar data is small dependence on atmospheric conditions and penetrating cloud cover. Use of imaging radars is independent of solar radiation so that imagery can be obtained at any time during day and night for continuous monitoring rice crops. The study used radar remote sensing data TerraSAR-X with dual polarisation HH&VV (band X, wavelength of 3.1 cm) and StripMap mode with very high spatial resolution (3 m) for monitoring and mapping distribution of rice-growing areas in Cho Moi district of An Giang province. The relationship between backscattering coefficient of TerraSAR-X images with the growth of rice plants was analysed in this research study. The backscattering coefficients of the two polarisations HH and VV change in the rice growing stages. Backscattering coefficient of HH polarisation is higher than VV polarisation from 20 days after sowing. In the last stage of rice crop, backscattering coefficients of HH and VV polarisation is close to each other and relatively stable. Especially, in the middle stage of crop, backscattering coefficient of VV was significantly reduced and then increased again, thus the polarisation ratio of HH/VV increased highly in this stage. Using the single-date HH&VV ratio image acquired in the middle of the crop and thresholding method, the distribution of rice-growing areas are mapped in order to support for crop monitoring and management precisely and detailed at district and commune levels. The study used dual-polarisation TerraSAR-X imagery for monitoring and mapping of rice provided good results when compared to statistics of rice area in Cho Moi District.