The MSS-UKMO Numerical Atmospheric-dispersion Modelling Environment (NAME) air quality and dispersion model runs for the Southeast Asia region are developed and operated by MSS in collaboration with the UK Met Office. It simulates the transport, dispersion, chemical transformation and eventual deposition of pollutants in the atmosphere. It provides 00 UTC smoke forecasts with a spatial resolution of 0.1° x 0.1° and a forecast range of 48 hours extendable to 72 hours.

Fire radiative power (FRP) products are obtained from CAMS-GFAS and applied over a high-resolution landcover and peatland map (Miettinen et al., 2016) to derive smoke emissions calibrated for the SEA region. Regional anthropogenic (REAS Version 3.2.1; Kurokawa and Ohara, 2020), biogenic (CAMS-GLOB-BIO; Granier et al., 2019; Sindelarova et al., 2014) and shipping (EDGAR Version 4.3; Janssens-Maenhout et al., 2012) emissions are also used.

The NAME model chemistry scheme includes the formation of secondary particulate matter and is used to generate realistic background pollutant levels compared to earlier system versions described in Hertwig et al. (2015). Aerosol optical depth (AOD) forecasts are computed based on PM2.5 aerosol optical properties derived from an empirical study of biomass burning smoke over Singapore by Lee et al. (2016).

The RSMC-VFSP Singapore multi-model ensemble (Baklanov et al., 2021; O’Neill et al., 2023) comprises one regional model, i.e., MSS-UKMO NAME (Singapore), and 5 global models, i.e., CAMS-ECMWF (Europe), JMA MASINGAR (Japan), NASA GEOS-5 (USA), NOAA GEFS-Aerosols (USA), and FMI SILAM (Finland). The forecast products include the ensemble and member predictions of smoke AOD and surface concentrations of PM10 and PM2.5.

The regional domain covers major burning areas and smoke transport pathways at a common grid resolution of 0.5° x 0.5°. The ensemble and member forecasts are provided at 3-hourly intervals up to 48 hours ahead. The ensemble has two products describing centrality (multi-model median and mean) and another two describing the spread (standard deviation and range of variation). 

A geo-referenced map of daily fire activity, depicted by fire hotspots detected by polar-orbiting satellites operated by the US National Oceanic and Atmospheric Administration (NOAA) is generated each day and included as an overlay on polar-orbiting satellite images. The depicted hotspots represent locations with possible active fires. An option for selecting the date allows the user to view past hotspot data.

Vegetation fires occurring in Southeast Asia are highly variable due to the combination of small farm holders’ land-clearing activities and wildfires (Reid et al., 2013), making forecast of fire activities in Southeast Asia highly uncertain. Satellite-based observations of fire activities may also be obstructed by cloud cover over SEA, and thus underestimated. Therefore, MSS-UKMO NAME model assumed persistence of fire activities for the forecast horizon.