ESTIMATION OF ATMOSPHERIC COLUMN AND NEAR SURFACE WATER VAPOR CONTENT USING THE RADIANCE VALUES OF MODIS M. Moradizadeh a, , M. Momeni b , M.R. Saradjian a a Remote Sensing Division, Centre of Excellence in Geomatics Eng. and Disaster Management, Surveying and Geomatics Engineering Dept., University College of Engineering, University of Tehran, Tehran, Iran (mina_moradizadeh@yahoo.com, sarajian@ut.ac.ir) b Surveying and Geomatics Engineering Dept., Faculty of Engineering, University of Esfahan, Esfahan, Iran - mehdimomeni@yahoo.com KEY WORDS: MODIS, Column water vapor, Near surface MMR water vapour, AIRS, Ratio technique, Transmissivity. ABSTRACT: One of the most important parameters in all surface-atmosphere interactions (e.g. energy fluxes between the ground and the atmosphere) is atmospheric water vapor. It is also an indicator among others to modeling the energy balance at the Earth’s surface. Total atmospheric water vapor content is an important parameter in some remote sensing applications especially land surface temperature (LST) estimation. As such, total atmospheric water vapor content and LST are used as key parameters for a variety of environmental studies and agricultural ecological applications. Estimation of an accurate LST requires the atmospheric water vapor content estimation. This study is concerned with retrieving total atmospheric water vapor content (W) using Moderate Resolution Imaging Spectrometer (MODIS). We have used a ratio technique to estimate the column water vapor based on MODIS data. However Atmospheric Infrared Sounder (AIRS) column water vapor and AIRS MMR near surface water vapor have been taken into account to calculate coefficients of the equation in the ratio technique. Then the accuracy of the results was examined using independent data set. It is concluded in this study that MODIS data is appropriate in mapping water vapor content as a suitable alternative to meteorological stations measurement data. In this paper, an existing operational algorithm is used to retrieve total atmospheric water vapor content from MODIS data using an LST independent approach. This paper offers a radiance based algorithm for retrieving total atmospheric water vapor content (W) using MODIS radiance data. As a new approach, AIRS data are taken as the reference data to calculate the coefficients of the equation of the method. AIRS is a facility instrument whose goal is to support climate research and improved weather forecasting. The AIRS instrument measures the distribution of water vapor in the atmosphere in three dimensional, globally, every day. 1. INTRODUCTION Meteorological solution to obtain atmospheric water vapor content consists of using radiosonde observations. Radiosonde observations are balloon based observation which cover a single profile from the land surface to about 30 km above the surface. The synoptic properties of the weather stations (which are the official for radiosonde measurements) and the point wise measurements of radiosonde data limit the applications. Remote sensing methods that allow us to estimate the atmospheric water vapor content have been developed in recent years. Studies in retrieving total atmospheric water vapor content (W) has been carried out using sensors such as the Along-Track Scanning Radiometer (ATSR), the Advanced Very High Resolution Radiometer (AVHRR), and the Moderate Resolution Imaging Spectrometer (MODIS). The radiance data have been split into two data sets. The first three data sets have been used to estimate the coefficient of equations and the fourth one has been used to test the coefficient. Validation of the total atmospheric water vapor has been done using a field data set. The data set contain near surface water vapor measured by weather stations. There are two main approaches for estimation of total atmospheric water vapor content using remote sensing data (Schroedter-Homscheidt and Drews, 2007). The first approach uses some regression based statistical relations which are based on brightness temperature of remote sensing thermal image pixels (Schroedter –Homscheidt and Drews, 2007). Also, since clouds are white and cooler than land surface, in this study, a simple method to detect cloudy pixels is using NDVI. In this method if 0 < NDVI , the pixel is cloudy and is eliminated (Ackerman, 1996). The second approach uses direct radiative transfer equation which has total atmospheric water vapor content implicitly. The radiative transfer equation gives the remote sensing sensor radiance as an addition of terms so that the total atmospheric water vapor content is implicit parameter in them (Schoroedter and Homscheidt, 2007). It is well studied that the second approach can reach to better accuracies because of its physical inheritances. However, the approach needs the knowledge of LST and surface thermal emissivity (Schroedter-Homscheidt and Drews, 2007). 2. DETERMINATION OF WATER VAPOR WITH MODIS DATA There are many approaches to estimate the water vapor from MODIS observations. They are, for example, the split-window difference of the thermal bands, the ratio technique, the regression slope, a look up table derived from radiative transfer model output (Kaufman and Gao 1992). 523