75 
SOIL MOISTURE ESTIMATION FROM RADARSAT -1, ASAR AND PALSAR DATA IN 
AGRICULTURAL FIELDS OF MENEMEN PLANE OF WESTERN TURKEY 
a Fusun Balik Sanli *, Yusuf Kurucu b , Mustafa Tolga Esetlili b , Saygin Abdikan 3 
a Yildiz Technical University, Faculty of Civil Engineering, Department of Geodesy and Photogrammetry Engineering , 
34349 Besiktas Istanbul, Turkey 
(fbalik, sabdikan)@yildiz.edu.tr 
b Ege University, Faculty of Agriculture, Department of Soil Science, 35100 Bomova Izmir, Turkey 
(yusuf.kurucu, tolga.esetlili)@ege.edu.tr 
Commission VII, WG VII/2 
KEY WORDS: RADARSAT-1, ASAR, PALSAR, Soil Moisture, Backscatter 
ABSTRACT: 
Due to the accelerating global warming, droughts which cause severe damages especially in the agriculture became a very recurrent 
phenomenon in all over the world. Monitoring the characteristics of soil moisture is very important in Turkey because the major 
impact of the global warming on our country appears to be climate changes. In this respect, drought has become a serious threat for 
the country where the agriculture is one of the major income sources. Therefore, monitoring of draughts has the highest priority 
among the other strategies. Although the sensitivity of microwaves towards the soil moisture is well understood, retrieving soil 
moisture with Synthetic Aperture radar (SAR) measurements still has difficulties due to the major impact of soil texture, surface 
roughness and vegetation cover. In this study, SAR data gathered by different sensors for the same area in closer dates were used to 
estimate the relative soil moisture. The relation between the ground soil moisture and the sigma nought/backscatter values of SAR 
images were investigated. Sigma nought values of C band HH polarized Radarsat Fine Beam image and C Band VV polarized 
ENVISAT (ASAR) images as well as backscatter values of an L band HH polarized ALOS (PALSAR) satellite images were used. 
RADARSAT, ASAR and PALSAR images were gathered on the 28th of May, 8th of June, and 10th of June in 2006 respectively for 
the alluvial lands of Menemen Town, Izmir. Ground soil moisture measurements taken using gravimetric methods showed a good 
agreement with the backscatter values of the images obtained from different types of SAR data. A comparison among the spatial 
distribution of retrieved soil moisture changes from SAR images was done. The correlations between the soil moisture content and 
backscattering of ASAR, RADARSAT-1 and PALSAR images were found 76%, 81% and 86 % respectively. Although the 
resolution of RADARSAT-1 fine beam image (6.25m x 6.25m) is closer to the resolution of PALSAR image (6.25m x 6.25m), 
PALSAR gives better correlation than RADARSAT-1 image. Although the resolution of RADARSAT-1 and PALSAR images is far 
more higher than that of the ASAR image (30m x 30m), the significance of the results produced is almost similar in such flat areas. 
1. INTRODUCTION 
Soils represent an important part of natural resources of Turkey. 
Soil management is an important element of sustainable 
agricultural development. The protection of soils has 
significance not only in the economy of the country but also in 
environment protection. Monitoring the characteristics of 
spatially and temporally distributed soil moisture is important 
due to the fact that the soil moisture controls the plant growth, 
the hydrological behaviour of the soil, and the ability of the soil 
to resist erosion, etc. Since Turkey has been located in the 
relatively drier parts of the earth, especially agricultural areas 
show vulnerability to reduced soil moisture. Parallel to the 
variations in the earth climate, the climate of our country also 
marks changes. Thus the precautions must be taken to decrease 
the effects of possible future droughts. 
The major difficulties in retrieving the soil moisture with SAR 
images are due to soil texture, surface roughness and vegetation 
cover. The amount of moisture stored in the upper soil layer 
changes the dielectric constant of the material and thus affects 
the SAR return. Because the dielectric constant for water is at 
least 10 times bigger than that of the dry soil, the presence of 
water in top few centimetres of bare soil can be detected in 
SAR imagery (Lillesand and Kiefer, 2000). The large difference 
between the dielectric constant of water and of dry soil at 
microwave frequencies is the main factor for soil moisture 
estimation using microwaves (Wang, 1980). Since the 1960s the 
radar remote sensing has been used operationally. Radar sensors 
transmit microwave energy to the earth surface and record 
backscatters reflected by the objects on the ground. The 
wavelengths of the energy used by these active remote sensing 
systems vary with frequencies between 0.3 GHz to 300 GHz 
roughly corresponding to wavelengths between 1mm and lm, 
with wavelengths between 0.5 and 50cm being widely utilized 
(Skidmore, A., 2003). For instance, the X-band within the range 
of 2.4-3.75 cm wavelength is usually reflected by the surfaces 
of the objects. The C-band falling into the wavelength range of 
3.75-7.5 cm could only reach to the parts near to the surface of 
the objects. The L-band which has the wavelength range of 15- 
* Corresponding author: Fusun Balik Sanli e-mail: fbalik@yildiz.edu.tr