Simulating Tropical Squall Lines Across Model Resolutions using a Scale-Aware Cold-Pool Gust-Front Parameterization

We implement the evaporation-driven cold-pool gust-front parameterization of Freitas et al. (2024) in the Model for Ocean-Land-Atmosphere Prediction (MONAN) and examine its performance together with revisions to convective scale awareness and entrainment. The parameterization is coupled to the Grell–Freitas–Li convection scheme. Four case studies over the Amazon basin, equatorial Africa, and the Maritime Continent span horizontal grid spacings from 60 to 3 km. We compare precipitation with GPM IMERG and MSWEP estimates and use ERA5 as an additional model-based reference. The results show that the cold-pool scheme significantly improves convection organization compared with the control runs. Global and regional experiments provide quantitative assessments of rainfall intensity, clustering, and forecast agreement. The new configuration produces more distinct precipitation bands and embedded cores and reduces widespread weak rainfall. It also better maintains heavy rainfall with increasing forecast lead time. Fractions skill scores improve in 160 of 180 evaluated combinations of threshold, lead window, and neighborhood width, with the clearest gains for moderate-to-heavy precipitation over broader neighborhoods. Object-based verification likewise shows improved detection of observed heavy-rainfall systems and smaller errors among matched systems, although a larger fraction of forecast objects remains unmatched. The benefits are therefore not uniform across verification measures. Increasing resolution strengthens local maxima but does not consistently correct errors in system location, continuity, and evolution. These results support the revised configuration for representing organized tropical precipitation while identifying remaining limitations in intensity and propagation.