Fast approximation for calculating deep convection cloud top heights from satellite brightness temperature

Due to fuel efficiency most air traffic takes place at the top of the troposphere, always avoiding deep convective clouds because of associated hazards such as severe turbulence, icing, hail, and lightning. If deep convection cloud tops are not very high, airplanes can sometimes fly above them, deviating less from their planned flight paths and causing less delays. Because of this, deep convection cloud top diagnosis and forecast is very important in aviation meteorology. One of the methods to estimate altitude of existing convective cloud tops is to compare infrared satellite brightness temperature with a calculated parcel curve temperature (parcel theory). The premise here is that we already have deep convective clouds so we are not interested in the full vertical temperature profile (sounding). To calculate the cloud top pressure, which is directly related to altitude in the standard atmosphere, we only need the parcel curve and measured satellite brightness temperature. Parcel curve is usually calculated iteratively from surface temperature and dewpoint, but this calculation can be computationally intensive for many points. In the first part of this work we will test different approximations of moist adiabat calculation and compare them to classical iterative method to see which one gives us acceptable error for estimating cloud top heights from satellite data. The second phase will implement the best approximation in operational forecaster environment.

We will also try to test some expected limitations of this approach, as calculation of parcel curve from only surface temperature and dewpoint is realistic only for diurnal deep moist convection, while in the case of elevated convection the most unstable layer temperature and dewpoint should be used.