1 A GLOBAL ANALYSIS OF PASSIVE MICROWAVE BRIGHTNESS TEMPERATURE DIURNAL CYCLE Zahra Sharifnezhad 1 , Hamid Norouzi 2 , Reginald Blake 2 , Emmanuel Gil 2 1 City College of New York, 160 Convent Avenue, New York, New York 10031, USA 2 New York City College of Technology, 300 Jay Street, Brooklyn, New York 11201, USA ABSTRACT Passive microwave brightness temperature (TB) radiometers are widely used to retrieve several atmospheric and surface parameters such as precipitation, soil moisture, freeze and thaw, water vapor, air temperature profile, and land surface emissivity. Since TBs are measured at different microwave frequencies with various instruments, incident angles, footprints, spatial resolutions, and radiometric characteristics, a combination of data from different microwave sensors could be inconsistent. For this reason, this study primarily uses the non- synchronous Global Precipitation Measurement (GPM) Microwave Imager (GMI) measurements to construct the diurnal cycle of TBs for each month. This diurnal pattern could be used as a point of reference to validate and calibrate the diurnal cycle of TBs from fusing the measurements of other sensors with daily fixed acquisition times (SSM/I, SSMIS, AMSR2, and etc.). The data from these sensors should also be merged and harmonized in order to build a comprehensive global diurnal cycle of passive microwave TBs. This highly frequent diurnal cycle will eventually help predict the emissivity estimation and potentially further advance the accurate prediction of the estimated time of the freeze/thaw (FT) transition states. Global comparison of TBs obtained from different sensors shows a moderate variance with a significant dependence on land cover type. The results of this study can enhance the temporal detection of freeze and thaw which is more helpful during the transition times when multiple FT changes may occur within a day. Index Terms: Brightness temperature, diurnal cycle, GMI, SSMI, AMSR2 1. INTRODUCTION Instantaneous measures of Passive microwave (PMW) brightness temperature (TB) have been used in a variety of applications to retrieve a number of atmospheric and surface parameters such as air temperature profile estimation, column water vapor abundance, precipitation rate, surface ocean wind speed, ocean salinity, soil moisture, freeze/thaw (FT) state detection, land surface temperature, inundation fraction and vegetation structure [1-5]. It can also be used in land-surface microwave emissivity retrievals as described by Prigent et al. 2015. The computational method is dependent on high temporal resolution TB which includes the component of cosmic background radiation [6]. Hence, a half-an- hourly diurnal cycle of the microwave TB will contribute to more accurately characterize emissivity values and other physical parameters retrievals. PMW sensors provide TB observations at different times and twice daily for sun-synchronous systems. As a result, an integration of data from several satellites is needed if a complete diurnal variation of TB is required. Since various instruments measure TBs at a large spectral range using a number of channels and two polarizations (horizontal and vertical) with incident angles, footprints, and radiometric characteristics [7], an amalgamation of TBs from different microwave sensors would not be necessarily consistent. However, once harmonized, they can provide a complete dataset to estimate TB diurnal cycle. In this work, observations from several PMW sensors are utilized to construct TB diurnal cycle at different channels. 2. DATA AND METHOD The coordination of satellite sensors that constitute the GPM constellation represents a step forward in the capability of the Earth observing system. GMI, the multispectral PMW instrument is the hub of this constellation, a to which all other constellation members are intercalibrated [1,8]. Moreover, the acquisition times of GMI change from day to day [9,10]; thus, the monthly shape (amplitude and phase) of the diurnal cycle could be obtained by integrating several days of measurements. Hence, a half-an-hourly diurnal cycle of TB is constructed using the GMI measurements in the first step. To build the diurnal cycle, a smoothing Spline interpolation method is fitted on all GMI TB observations within each/a month. In this method, the data is fitted to a set of Spline basis functions with a reduced set of knots, typically by least squares.