Abstract
Ph.D. (Environmental Science)
A number of studies to detect effects of atmospheric acidic deposition on regional environments
have been conducted in South Africa in the past, without finding any clear evidence of adverse
effects. Despite these studies, scepticism remained that acidic deposition could accumulate on the
decadal scale and reach a point where the natural buffering capacities of soils and water bodies
would be exceeded. Against this background, this study was conceived to make direct
measurements of atmospheric concentrations of the acidifying gases, over as an extensive area as
possible, to provide an objective basis to the question of whether or not acidic deposition pose
immediate or long-term threats to regional ecosystems. As an integral study, the research strategy
was designed to address the question as to whether ozone, as a secondary pollutant, would exceed
the tolerance of the same ecosystems.
A passive monitoring network was devised to measure monthly mean atmospheric concentrations
of three trace gas species, SO2, NO2 and O3. The network comprised of 37 monitoring sites at
remote locations over the northern and eastern portions of South Africa, at 1º grid intervals (0.5 º
for several sites). Trace gases were sampled monthly over two complete annual cycles at each site,
using passive diffusive samplers exposed for approximately 30 days. Samples were chemically
analysed at an internationally accredited laboratory (Atmospheric Chemistry Research Group,
North-West University) and mean monthly results were recorded.
The collated database of trace gas concentrations enabled assessment in terms of standards, time
and spatial distributions. Concentrations were evaluated in a critical levels assessment against
several sets of South African and international standards and guidelines. The dry deposition rates
were calculated from measured ambient trace gas concentrations, using an inferential model. For
estimates of wet deposition, long-term acidic rainfall measurements were multiplied by the
cumulative precipitation within the sampling period amount for each project site. These total dry
and wet deposition estimates (for sulphur and nitrogen compounds), and concentrations of ozone
were compared with results from two regional-scale studies of modelled concentration and
deposition. The total acidic deposition estimates were adjusted with mitigating base cation
deposition estimates, dry and wet, and the derived net acidity loads were subtracted from soils acid
buffering (sensitivity) capacity loads to determine critical loads exceedance.
Two ancillary tasks were undertaken, which contribute to the central aim. The first accessed the
most recent measured regional lighting strokes data and adjusted it for intra/inter-cloud strokes to
estimate the lightning produced NOx (LNOx) budget. High and a low lightning NOx budget
estimates were then compared with three relevant anthropogenic emission inventories compiled for
the Highveld region. The second task obtained regional SO2, NO2 and O3 remote sensing
information and compared them with ground level concentrations from the passive sampling
network. Remote sensing total column densities were converted and partitioned through
atmospheric pressure averaging equations to derive tropospheric boundary concentrations
appropriate for comparisons.
vi
The main findings were:
• Concentration distributions for acidic gases SO2 and NO2 show prevailing high
concentrations over the industrial Highveld. The areas downwind show substantially
decreased concentrations, while remote areas show very low concentrations. Ozone
concentrations are spatially uniform, with slightly increased concentrations located away
from the central Highveld pollution source area, both north and south. However, ozone
distributions and origins could not be directly related to the Highveld industrial emissions.
No health-based exceedance (of long-term standards) was recorded for SO2, NO2 or O3.
Three areas of critical exceedances were found for SO2, one in the industrial Highveld for
sensitive lichen and semi-natural and forest vegetation; and two downwind, for sensitive
lichen only. No ecosystem exceedance was established of NO2 and O3 critical levels.
• Acidic deposition distributions showed a direct relationship to high atmospheric
concentration distributions, with exceptions for those areas which had high annual rainfall,
and for coastal and escarpment areas where wet deposition prevailed. No concurrent regional
modelling had been performed by other parties. Two non-concurrent regional modelling
studies were selected for comparison of measured and modelling results. Substantial
disparities were identified between the respective results and, although general patterns were
followed, the modelling for dry and wet acidic compounds swayed from overestimates to
underestimates for sulphur and nitrogen species. In contrast, modelling of ozone mainly
overestimated measured concentrations. In terms of spatial distributions, ozone modelling
matched a wide band of elevated concentrations in northern area of South Africa (Limpopo
Province).
• Derived net acidic deposition loads were matched to critical loads from soil types in the
region. This revealed areas with high deposition and highly sensitive local soils within the
greater industrial Highveld with critical load exceedance. An area north-west of the central
Highveld showed exceedance of the highest critical load, while several smaller areas
downwind to the south-east showed exceedances of lower critical loads only.
• The lightning NOx budget estimation showed that lightning NOx is a significant contributor
to overall regional NOx budget and a major natural contributor. Depending on lightning
count assumptions applied, annual LNOx production varied from a 1/10th to 1/3rd of the NOx
anthropogenic trace gases emissions, taken from the most comprehensive of three recent
inventories. However, when seasonal lightning budgets were examined, no distinct seasonal
increases in ground-level NOx were found in areas of high lightning frequency, suggesting
that much of LNOx is generated and transported above the boundary layer.
• Comparisons of remote sensing concentrations results for SO2 and O3 have not revealed very
good spatial and temporal agreement, although both showed agreement within an order of
magnitude between measured and remotely sensed concentrations. NO2 agreement was much
better in both spatial and temporal comparisons. The mean ratio of remotely sensed to
measured NOx is 0.77, an encouraging result given the broad assumptions made in deriving
the comparison. Disparities in the horizontal distributions between ground level and satellite
products for NOx indicate a need for better information on the vertical distributions of NOx
vii
and other trace gases – the assumption of a uniformly mixed boundary layer is not robust
enough for such comparisons.
From findings of all individual studies an overall answer to the central question is reached that
pollution from acidic gases and their compounds is not a current or medium term threat to regional
ecosystems (beyond the central pollution source area), at current rates of emission. Results from the
control site, positioned specifically within the well-researched industrial Highveld, confirmed the
known situation in this sub-region, while serving well to put in perspective the entire region. Ozone
distributions indicate uniform regional distributions, with most monthly means below 30 ppb, and
highest single month 43 ppb. Now that this project has established a detailed measured set of acidic
trace gas concentrations, limiting factors in determining critical loads are (i) more detailed maps of
soil sensitivity (acidic buffering capacity); and (ii) base cation deposition measurements, including
dry deposition of dust, and wet deposition, both at more sites and over similar periods used in the
current study for the acid components.