Publications·November 30, 2024
This UNESCAP working paper reviews how Viet Nam can use satellite and ground-based remote sensing to strengthen air-quality management. It sets the pollution context (levels, sources, health burden), maps institutions and monitoring networks, summarizes what EO (Earth Observation) is already being used, documents the roll-out and current status of Pandora spectrometers and the GEMS geostationary mission, and offers practical recommendations to move from research pilots to operational, decision-ready products. Headline context: annual mean PM₂.₅ exposure ≈ 20 µg/m³ (≈4× the WHO guideline), with ~37,457 deaths in 2019 attributable to fine particles; 2023 PM₂.₅ levels in the Northern Region exceeded national standards by 1.4–2.1×.
Pollution sources, trends, and institutions
The report attributes urban/industrial air quality problems primarily to traffic and energy: traffic emissions account for ≈70% of air pollution, 85% of CO and 95% of VOCs; coal use in power and industry is growing and projected to increase mortality (e.g., ~188 excess deaths per million by 2030 under continued coal expansion). A multi-city analysis (2010–2018) shows declines in most criteria pollutants in Hà Nội except SO₂; seasonal biomass burning adds significant PM and CO, with fire hotspots concentrated in the northwest.
Who does what. The Ministry of Natural Resources and Environment (MONRE) leads; within it, the Viet Nam Environment Administration (VEA) and the Center for Environmental Monitoring (CEM) run national monitoring and data systems. At the local level, DONREs (provincial Departments of Natural Resources and Environment) implement and report; the local-government hierarchy (province → district → commune) provides the “last-mile” for management, budgeting and enforcement.
Ground monitoring — what exists today
Viet Nam operates parallel automatic and manual networks:
Automatic (national): 37 standard stations + 4 sensor stations managed centrally;
Provincial: 85 standard + 42 sensor stations under DONREs;
Manual: 94 locations across 25 provinces.
Other contributors include US AirNow (Hà Nội) and GIZ (Hồ Chí Minh City) pilots with low-cost sensors. The Environmental Monitoring Portal publishes hourly AQI for PM₂.₅, PM₁₀, NO₂, SO₂, O₃, CO at dozens of stations, but spatial gaps remain — especially outside major cities — where satellite products can fill in.
What remote sensing is already doing in Viet Nam
Satellites & products. Researchers combine MODIS/VIIRS AOD, TROPOMI/OMI/GOME-2/SCIAMACHY trace-gas columns (NO₂, SO₂, HCHO), and reanalyses (e.g., MERRA-2) with machine-learning/statistical downscaling to estimate surface PM₂.₅ and NO₂ at city-to-national scales (e.g., 1×1 km NO₂ maps for 2019–2021; 3×3 km daily PM₂.₅ for 2012–2020). Multiple studies document long-range transport into Hà Nội during spring burning seasons. While the national publication base is growing, EO products remain fragmented across academic projects and are not consistently open or operationalized by government portals.
National EO capacity. The Viet Nam National Space Centre (VNSC) has flown PicoDragon (2013) and MicroDragon (2018), and earlier VNREDSat-1 provided 2.5 m optical imagery — assets that demonstrate domestic EO engineering and operations.
Fires & advisories. Since 2007, the Forest Protection Department has run an operational FireWatch system using MODIS/AVHRR direct-broadcast for real-time hotspots, archiving, and statistics; Global Forest Watch and the FIRED dataset complement national analytics. Seasonal peak: late Jan–May.
Regional cooperation. Viet Nam participates in SERVIR-Mekong (near-real-time satellite layers + 3-day forecasts) and ASEAN haze mechanisms (MSC-Mekong), which set hotspot-reduction targets (e.g., 35% in 2024, 40% in 2025 from 2004 baseline).
Pandora & GEMS — status and early lessons
Pandora (Pandonia/Pandora Asia Network). Three Pandora spectrometer sites are identified:
Hà Nội — Northern Center for Environmental Monitoring (PAN-226);
Đà Nẵng — Middle & Highland Center (PAN-222);
Southern Center (MONRE Complex) — PAN-223.
The report notes installation/commissioning steps and indicates the system is operational with additional site completion planned in late-2023. Pandora provides high-cadence column measurements (NO₂, HCHO, O₃, SO₂) critical for GEMS validation, diurnal profiling and local vs. transported diagnostics; success depends on QA/QC, uptime, tracker maintenance, spares, and co-location with reference PM/criteria-gas monitors.
GEMS (geostationary). Viet Nam has trained analysts on GEMS L2 ingestion (capacity-building in 2022, Korea/Thailand). Early applications compare GEMS AOD with automatic stations and assess feasibility for assessment, forecasting and early warning. Initial results show useful long-range transport signals but highlight needs for finer resolution, routine bias-correction, and harmonization with ground data and meteorology. A recent modelling study (regional) found GEMS-informed NO₂ emissions updates significantly improved surface NO₂ skill versus inventories alone — though overestimation of NO₂ columns in some Southeast Asian cities (incl. Hà Nội/HCMC) signals the need for continuous validation.
What remote sensing can close — priority knowledge gaps
Unmonitored areas. Use fused AOD + CTM (e.g., GEOS-Chem) with GWR/ML bias-correction to produce monthly/annual PM₂.₅ and NO₂/SO₂/O₃ surfaces for all provinces; build hotspot and trend maps for policy reviews.
SLCPs & precursors. Operationalize NO₂, SO₂, HCHO, O₃ mapping with GEMS/Pandora/TROPOMI for source diagnostics and chemistry, complementing climate actions.
Long-term trend & policy evaluation. Use EO+ground series to evaluate vehicle/fuel standards, industrial controls, burning bans, and city action plans; publish policy-effect dashboards.
Transboundary transport & airsheds. Combine columns/AOD with trajectory modelling to delineate airsheds and quantify cross-border episodes to inform ASEAN cooperation.
Equity & health. Merge high-resolution pollution fields with population/vulnerability layers to target interventions; link to composition work (e.g., MAIA) to prioritise the most toxic aerosol components.
Fire emissions. Move beyond hotspot counts to burn-area + FRP-based inventories by fuel/land-cover to evaluate prevention and agricultural-burn policies.
Recommendations — what to do next
Make EO part of the plumbing. Build a national EO–AQ platform to automate ingest of GEMS/TROPOMI/OMI, Pandora, AQMN (auto + manual), and MODIS/VIIRS fires; publish near-real-time maps, archives and open APIs with uncertainty metadata and bilingual documentation.
Harden Pandora operations. Standardize SOPs/QA-QC, co-locate with reference PM/criteria gases, ensure spares/power/tracker maintenance, and report uptime/flags transparently.
Operationalize GEMS workflows. Bias-correct with Pandora/AERONET/ground; create episode tools (straw-burn plumes, local vs. transported attribution, volcanic SO₂); tie products to alerts, enforcement and planning.
Invest in people. Continue large-scale training (EO ingest, ML bias-correction, WRF/CTM, nowcasting/forecasting), with a help-desk and shared code repositories; clarify institutional roles across MONRE/VEA/CEM/DONREs and VNSC.
Policy integration. Update the National Plan on Air Quality Management and related laws to name specific EO products and indicators; align with ASEAN haze targets and leverage EO for implementing the “25 Clean Air Measures for Asia and the Pacific” (e-mobility, renewables, dust control, post-combustion controls).
Key words
Viet Nam; PM₂.₅ exposure; traffic emissions; coal power; remote sensing; GEMS; Pandora; AOD-based PM mapping; NO₂/SO₂/HCHO/O₃; biomass burning & FRP/burn-area; airsheds & transboundary transport; MONRE/VEA/CEM/DONRE; SERVIR-Mekong; FireWatch.