Multi-frequency microwave radiometer observations are used for estimating vapor and temperature profiles, and also liquid water path. Temperature and vapor profiles in the boundary layer determine the extent of the cloud profile and its microphysical structure. Temperature inversion and moisture layer parameters might be particularly useful for estimating cloud formation and its growth. One of the retrieval techniques, namely, the neural network-based technique is based on a set of simulated brightness temperature data for a specified radiosonde data for estimation. Application of this method at various geographical locations is limited by the brightness temperature training dataset generated by the radiosonde data.

In this study, we investigate a physically based method for retrieving a temperature profile using ground based radiometric measurements. A parameterized radiation transfer model is used for quantifying the down-welling microwave brightness temperatures in 12 frequencies in K and V bands. The sensitivity of the down-welling brightness temperatures to Liquid Water Path (LWP), total water vapor, amount of temperature inversion, and height of the inversion layer are studied. A sensitivity study characterizing temperature profile with only a limited number of parameters such as lapse rate, amount of temperature inversion and height of the inversion layer might allow us to construct a physically based retrieval method.