Temperature variations can have significant effects on guide wave propagation, and therefore can increase the detection uncertainty of the structural health monitoring (SHM) system. The effect of this variation has been investigated for detecting impact damage in carbon fiber reinforced composites. Surface instrumented piezoelectric wafers are used in this work to detect the low velocity impact damage in a braided carbon fiber reinforced composite plate over a temperature range of 0°C to 60°C. The effects of temperature variation and thermal cycles on guided wave propagations in healthy composite plates are characterized first. The information is then used to develop a compensation algorithm to minimize the thermal effects on detection. A statistical damage detection and quantification approach is developed using and Auto-Regressive with eXogenous (ARX) model and local outlier analysis. The developed methodology is experimentally validated using the pitch-catch guided wave testing approach within the temperature range.