TL;DR
This study presents the first global, multi-year (2013-2020) map of biogenic isoprene emissions directly inferred from satellite observations of isoprene itself, rather than its indirect proxies. This new dataset offers a more robust constraint on atmospheric models, revealing that extreme climate events like El Niño and heatwaves significantly boost isoprene emissions, particularly in the Amazon.
What Problem Does This Paper Address?
Isoprene is the most abundant volatile organic compound released by plants, playing a crucial role in atmospheric chemistry, air quality, and climate. However, accurately quantifying its emissions has been challenging. Traditional "top-down" methods often rely on formaldehyde (HCHO) as an indirect indicator of isoprene, but HCHO is produced from many sources, leading to uncertainties. While recent satellite instruments can now directly measure isoprene in the atmosphere, global, long-term emission estimates derived from these direct observations were lacking. This paper aims to fill that gap by providing a comprehensive, multi-year global emission dataset directly constrained by satellite isoprene measurements.
Methods
The researchers employed a "finite-difference mass balance" (FDMB) inversion framework. This method uses a chemistry-transport model (LMDZ-INCA) to simulate atmospheric conditions. They then assimilated satellite-derived isoprene column concentrations into this model. The inversion process adjusts the model's simulated isoprene emissions until they best match the satellite observations. This approach directly constrains isoprene emissions, bypassing the uncertainties associated with using HCHO as a proxy. The model itself is coupled with the ORCHIDEE land surface model, which provides prior estimates of biogenic emissions based on factors like leaf area index and temperature.
Datasets
- CrIS isoprene columns: Retrieved from the Cross-track Infrared Sounder (CrIS) satellite, these are the primary observations used for the direct inversion.
- TROPOMI HCHO columns: Used for evaluating the model's performance in simulating HCHO, an indirect proxy for isoprene.
- OMPS HCHO columns: Another satellite-based HCHO dataset used for model evaluation.
- Pandonia Global Network (PGN) ground-based HCHO: Used for independent validation of the model's HCHO simulations.
- ERA5 reanalysis data: Provided meteorological variables like temperature, precipitation, and radiation.
- ESA Land Cover Climate Change Initiative (Land_Cover_cci): Provided land cover data.
- CEDS (Community Emissions Data System) and GFED (Global Fire Emissions Database): Used for anthropogenic and fire emissions, respectively.
The paper introduces a new global, monthly biogenic isoprene emission dataset for 2013-2020.
Key Insights and Ideas
- Direct isoprene observation is powerful: Using satellite measurements of isoprene itself, rather than its proxies like HCHO, provides a significantly more direct and robust constraint on emission estimates.
- Climate extremes drive emissions: Major climate events, specifically the 2015-2016 El Niño and the 2019-2020 heatwaves, are clearly linked to significant spikes in global isoprene emissions, highlighting the strong influence of temperature anomalies.
- Amazon is a key player in variability: The Amazon basin emerges as the largest contributor to year-to-year fluctuations in global isoprene emissions, underscoring its importance in the global carbon cycle and climate.
- Northern Hemisphere seasonality dominates: The pronounced seasonal cycle of isoprene emissions is largely driven by the Northern Hemisphere's strong temperature and vegetation seasonality.
- Bottom-up inventories may overestimate Oceania: The study suggests that some existing bottom-up inventories might overestimate isoprene emissions in regions like Oceania, as their seasonal patterns differ significantly from the satellite-constrained results.
Key Results
- Global average emissions: The study estimates global biogenic isoprene emissions to be 456 ± 238 Tg C yr⁻¹ over 2013–2020.
- Interannual variability: Emissions varied by an average of 14 Tg C yr⁻¹ (1-sigma standard deviation) annually.
- Emission peaks: Two notable emission peaks were observed in 2015–2016 (456 Tg C yr⁻¹) and 2019–2020 (478 Tg C yr⁻¹), coinciding with major climate events.
- Regional variability driver: The Amazon contributes 22.3% of the global interannual variance in isoprene emissions.
- Model performance improvement: The LMDZ-INCA model simulations using the new emission estimates showed improved spatial agreement and reduced biases compared to satellite HCHO products and ground-based measurements.
- Seasonal cycle: A pronounced peak in emissions occurs during July-September (JAS), with a minimum in December-February (DJF), driven by Northern Hemisphere seasonality.
Limitations and Future Work
- Incomplete satellite coverage: CrIS observations have limited coverage at high latitudes (north of 60° N), where emissions were kept at prior values. While this has a minor impact on global totals currently, future warming may increase the importance of these regions.
- Model-observation uncertainties: The ANN retrieval for isoprene lacks scene-specific vertical sensitivity, and the LMDZ-INCA model has inherent uncertainties in its chemistry and transport parameterizations.
- Linearity assumption: The FDMB inversion assumes a localized linear relationship between emissions and atmospheric concentrations, which may not hold true in all regions, particularly in high-isoprene, low-NOx environments like the Amazon.
- Future work: The authors suggest adopting joint NOx-isoprene inversions or iterative schemes to better capture complex chemical couplings. Extending the analysis to multi-decadal periods is needed to disentangle short-term climate variability from long-term drivers like CO2 and land-use change. Further investigation into the combined impacts of multiple environmental stressors (warming, deforestation, CO2, extreme events) on future isoprene emissions is crucial.
Summary updated