EGU24-8299, updated on 08 Mar 2024
https://doi.org/10.5194/egusphere-egu24-8299
EGU General Assembly 2024
© Author(s) 2024. This work is distributed under
the Creative Commons Attribution 4.0 License.

Using satellite observations to examine the role of land cover and fires in driving atmospheric composition over the southern Amazon

Emma Sands1, Richard Pope2,3, Ruth Doherty1, Fiona O'Connor4,5, Chris Wilson2,3, and Hugh Pumphrey1
Emma Sands et al.
  • 1School of Geosciences, University of Edinburgh, Edinburgh, United Kingdom
  • 2School of Earth and Environment, University of Leeds, Leeds, United Kingdom
  • 3National Centre for Earth Observation, University of Leeds, Leeds, United Kingdom
  • 4Met Office, Exeter, United Kingdom
  • 5Department of Mathematics and Statistics, Global Systems Institute, University of Exeter, Exeter, United Kingdom

We exploit satellite datasets of spatio-temporal distributions of atmospheric composition for the rainforest and savanna region on the southern boundary of the Amazon to understand how its emissions of biogenic volatile organic compounds (BVOCs) and local pyrogenic emissions impact the atmosphere. In particular, we explore the relationship between land cover change, considering vegetation type (e.g. broadleaf forest, savanna, grassland) and Leaf Area Index (LAI), and burned area and atmospheric composition. In this study, we investigate these relationships over the southern Amazon for the period 2001-2019, focussing on seasonal and spatial patterns. We utilise data for five chemical species: total column isoprene (TCC5H8), total column methanol (TCCH3OH), tropospheric column nitrogen dioxide (TCNO2), total column carbon monoxide (TCCO) and total column formaldehyde (TCHCHO), as well as aerosol optical depth (AOD).

We find burned area approximately delineates the areas of change in dominant vegetation cover type over time.  Robust relationships were found between TCC5H8 and forest cover, and TCNO­2­ and burned area. Here, we find that TCC5H8 linearly increases by 1 × 1014 molecules cm-2 with an increase of 1 percentage point in broadleaf forest cover. This is equivalent to densely forested regions having column isoprene ­­values four times greater than those with no­ forest cover. There is a strong power law relationship between TCNO2 and burned area. Overall, there is a larger increase in TCNO2 in regions of lower, though still substantial, biomass burning (i.e. potentially new regions of burning/deforestation). These relationships highlight the relatively short lifetimes of the two species such that their spatial extent is largely confined to their emission source regions.

Conversely, TCHCHO, TCCO and AOD reach maximum values for high broadleaf forest coverage and high burned areas, suggesting a mixed influence of both biogenic and pyrogenic sources, likely due to the longer lifetimes of these species and aerosols, allowing them to mix and be transported further from their emissions sources. Broadleaf forest cover and burned area do not appear to have a substantial impact on methanol, which is elevated over a region of savanna and grasslands in the northeast of the study region.

The results highlight the potential for air quality impacts from the biogenic and pyrogenic emissions and their interactions that differ seasonally and regionally, and illustrates how land cover and land use change exerts a strong control on isoprene and nitrogen dioxide concentrations over remote regions.

How to cite: Sands, E., Pope, R., Doherty, R., O'Connor, F., Wilson, C., and Pumphrey, H.: Using satellite observations to examine the role of land cover and fires in driving atmospheric composition over the southern Amazon, EGU General Assembly 2024, Vienna, Austria, 14–19 Apr 2024, EGU24-8299, https://doi.org/10.5194/egusphere-egu24-8299, 2024.