Accurate attitude information of satellites is a prerequisite for many space missions requiring some pointing capabilities. A radiation-based (RB) thermopile measurement package has recently been proposed for satellite attitude estimation with potential cost and power benefits. However, previous related researches have not considered all modes of the thermal communication, and have specifically neglected the conduction effect on and between the satellite surfaces as a simplifying assumption to initially develop the innovative measurement equation. As this assumption is practically difficult to implement for contemporary satellites operating in near Earth space environment, the present work complements the effect of conduction terms both in the measurement model as well as in relations pertaining to the solar RB attitude knowledge. In addition, another simplifying assumption of constant Earth albedo factor (AF) is also removed. Moreover, the present study simultaneously estimates the satellite attitude as well as its surface’s thermal characteristics (TC). The stochastic estimation algorithm used in the current study is based on the unscented Kalman filter (UKF) for attitude estimation via the new proposed sensor pack and measurement model. Finally, the study is augmented with a thorough sensitivity analysis against satellite internal heat radiation, environmental and initial conditions that have been verified via Monte Carlo simulations (MCS) to assess the merits of the proposed modified thermal model for RB concurrent satellite attitude and TC estimations.