What is the two-temperature method for calibrating a noise figure measurement system?

The two-temperature method for calibrating a noise figure measurement system uses two known noise sources at different physical temperatures (a hot load at temperature T_hot and a cold load at temperature T_cold) to establish the relationship between input noise power and output power reading, providing absolute noise temperature calibration of the measuring receiver. The method works by: connecting the hot load (typically an ambient-temperature termination at T_hot approximately 290-300 K) to the receiver input and measuring the output power P_hot, then connecting the cold load (a cryogenic termination immersed in liquid nitrogen at T_cold = 77 K, or liquid helium at 4.2 K) and measuring P_cold. The receiver noise temperature is calculated from: T_receiver = (T_hot - Y x T_cold) / (Y - 1), where Y = P_hot / P_cold (the Y-factor). The gain is G = (P_hot - P_cold) / (k x BW x (T_hot - T_cold)). The advantage over the standard ENR noise source method is higher accuracy: the two physical temperatures are precisely known (liquid nitrogen at 77.36 K is an SI-traceable standard), whereas ENR noise sources have calibration uncertainty of +/- 0.1-0.2 dB that directly limits NF measurement accuracy. The two-temperature method achieves absolute noise temperature accuracy of +/- 1-3 K, which is essential for radio astronomy receivers where T_receiver is only 3-10 K and even 1 K error represents a 10-30% measurement uncertainty.