Reducing Moist-Adiabatic Calculation Costs Using Lookup Tables

Tracking the temperature of an ascending or descending saturated parcel is complicated but essential for evaluating ubiquitous parameters like CAPE and CINH. Methods for calculating the final parcel temperature based on empirical functions (e.g., Davies-Jones 2008) are widely used. However, the cost of these operations can become substantial when model data containing thousands or millions of vertical profiles are processed, particularly when parameters are desired that require tracking multiple parcels per profile (e.g., effective-layer shear).

In some instances, lookup tables have been implemented to avoid the cost of directly calculating saturated parcel temperature changes. Here, an updated method is proposed in which fully-converged solutions of the empirical equation from Bolton (1980) are stored for a range of LCL temperatures and pressures defining meteorologically-relevant parcels. The table is then read in at the start of post-processing and simple interpolation of the tabulated moist adiabat temperatures is used to determine the final temperature for a given parcel. Large model output fields from the 2019 NOAA Hazardous Weather Testbed Spring Forecasting Experiment are processed to compare the accuracy and speed of this approach to that of the Davies-Jones (2008) empirical function method and the θ_e-based table method currently used in the NCAR RIP post-processing program.