Physical processes of fog in the Brazilian Northeast: Forecast by PAFOG and FogVIS models

Abstract

A specific model for low-visibility forecasting in the Brazilian Northeast (BNE) has not been developed; therefore, the German Parameterized Fog (PAFOG) model was adapted for the region. Additionally, Fog Visibility (FogVIS), a simple equation-based tool, was developed and requires further testing. From 2008 to 2020, Meteorological Aerodrome Report and Terminal Aerodrome Forecast surface data were collected via the Meteorology Network of the Brazilian Air Force Application Programming Interface, identifying 218 fog events across three airports: Maceió (32 events), Recife (1 event), and Campina Grande (185 events). GOES satellite images were accessed from the Center for Weather Forecasting and Climate Studies database, and synoptic and thermodynamic analyses were performed using ERA5 reanalysis data. Humidity from nearby water sources (lagoon for Maceió, dam for Campina Grande) was a primary factor in fog formation. PAFOG demonstrated strong predictive performance for Maceió and Recife’s single brief events, especially in 12-h forecasts, particularly when fog events were preceded or followed by mist or light rain. In contrast, FogVIS often aligned closely with the observed visibility range and provided complementary results 18 hours in advance for Campina Grande’s events, which were more intense but less associated with rain or mist, and also showed higher Fog Stability Index results. Both models demonstrated efficiency, with PAFOG excelling in Maceió and FogVIS in Campina Grande, highlighting the applicability and accuracy of both models in predicting fog for the BNE.