first, stability is bought over years, not necessarily days, and for bound not over short periods of time like sunrise and sunset and clouds. 2nd, matlab Tangaroa test has a minimum of two and perhaps more explanations: matlab one which you describe, and matlab augment in vaporization brought on by increased LWIR; where and when water is evaporating, engineering small increment of LWIR raises matlab evaporation rate without affecting matlab rate of heat diffusion from below. Matthew Marler,“first, balance is got over years, not always days”It depends upon matlab size scale thermal inertia. People speak of micron layers penetrable by IR; undoubtedly they need to stability in seconds or less. In matlab top 5 cm used for skin final result, engineering 1 W imbalance will raise matlab temperature 1°C in about six hours. etc.