United States
Environmental Protection
Agency
Environmental Research
Laboratory
Corvallis OR 97333
Research and Development
EPA/600/S3-89/030 July 1989
v>EPA Project Summary
Regional Analysis of
Wet Deposition for
Effects Research
Richard Vong, Steven Cline, Gregory Reams, Joseph Bernert, Donald Charles,
James Gibson, Timothy Haas, John Moore, Rudolf Husar, Anthony Olsen,
Jeanne Simpson, and Steven Seilkop
The basis for regional charac-
terization and analysis of precip-
itation amount, concentration, and
deposition Is Investigated. Key issues
in spatial analysis are the data
selection, data compositing, the
interpolation technique used, and the
uncertainty of the results. Sources of
data on precipitation amount and
chemical composition are presented
along with procedures for screening
the chemical data. A review of recent
work reveals that different scientists
select different data sets and that
data selection plays an important role
in the resulting maps.
Important issues In data prepro-
cessing Include temporal resolution,
data stratification into geographic
regions, and choosing between direct
and indirect methods for interpo-
lating wet deposition. The "indirect
method" involves interpolating prec-
ipitation amount and concentration
separately and using their product for
wet deposition maps. The Indirect
method is recommended because it
allows the use of more spatially
dense precipitation amount data sets.
Limited experimental evidence
demonstrates no spatial correlation
between that precipitation amount
and concentration, a necessary con-
dition for the use of the indirect
method. It Is recommended that
further investigation of the degree of
independence of precipitation
amount and concentration across
space be performed.
There are many methods to weight
near and distant data for estimating
data at a non-monitored site. The
geostatlstical technique, krigfng, is
discussed In detail to allow other
researchers the benefit of previous
applications to precipitation chem-
istry. Different Interpolation tech-
niques may produce maps that are
similar but estimation variances that
are different, or absent Procedures
for generating and checking uncer-
tainty estimates are discussed.
This Protect Summary was devel-
oped by EPA's Environmental Re-
search Laboratory, Corvallis, OR, to
announce key findings of the research
project that is fully documented in a
separate report of the same title (see
Project Report ordering information at
back).
Summary and Conclusions
Recent investigations, fully referenced
by the authors in the final report, reveal
that eastern U.S. forests are declining
and that some eastern lakes have
become acidic. One possible hypothesis
is that acidic deposition represents a
stress to these ecosystems that, when
combined with natural stresses, has
caused these changes. It has been
suggested that three conditions must be
satisfied to prove causation: (1) a
mechanism, (2) a dose-response relation-
-------
ship, and (3) spatial and temporal consis-
tency. To examine spatial consistency
one constructs maps of potential stresses
and damage. If the gradients in the wet
deposition of certain chemical species
and forest damage (or lakewater acidity)
are similar one requirement for a proof of
causality has been met, namely "spatial
consistency."
Maps are very useful for visualizing
spatially oriented data. Regional esti-
mates expressed as isopleth maps can
reveal the magnitude and extent of acidic
deposition and locate areas of high or low
deposition. Spatial interpolation is used to
generate maps of wet deposition and to
estimate data for non-monitored loca-
tions. Interpolated values for wet deposi-
tion are used as inputs to aquatic and
terrestrial ecosystems in the absence of
measurements. This report reviews
issues related to interpolating data and
charting patterns in precipitation chem-
istry and wet deposition.
The chemical species of likely interest
to terrestrial researchers and limnologists
include those which control the acidity of
the precipitation or induce/neutralize
acidity in receiving systems such as the
concentration and/or deposition of NOa',
S04 = ,H + ,Ca* YandNH/.
f-T and NH4* concentration or wet
deposition are of interest due to the
potential acidifying effects of these ions
on foliage and soils. S04 = , and N03"
concentrations are of interest because
they usually derive from anthropogenic
emission sources (at Northern Hemi-
sphere continental monitoring sites) and
because they may be involved in cation
leaching from soils or loss of acid
neutralizing capacity from lakes. Cation
deposition might be of interest because
of added buffering to soils or lakes. NO3
and K* can serve as nutrients for plants.
Either wet deposition or precipitation of
chemical concentrations may be relevant
depending on the ecological effect of
interest. For example, foliar leaching in
spruce needles or lake chemistry in
watersheds with thin soils might be
related to precipitation concentrations
while spit buffering processes might
respond to wet deposition. Dry deposition
or cloudwater interception can contribute
substantial chemical inputs to terrestrial
ecosystems, especially at high elevations
in the eastern U.S.A. Only wet deposition
(precipitation) is considered because it is
the best understood and most intensively
monitored of the three deposition path-
ways.
A number of approaches for strength-
ening any spatial analysis of precipitation
chemistry are presented. The goal is to
ensure that future investigators will
benefit from the experiences discussed
herein and that they will document key
portions of their analyses to permit
evaluation by peers.
When performing a regional analysis of
precipitation chemistry four key issues
are: the data selection, data compositing,
the interpolation technique, and the
uncertainty of the results. Using National
Atmospheric Deposition Program
(NADP/NTN) and Canadian precipitation
chemistry data (at a minimum), screening
the chemical data, and using supple-
mental National Weather Service (NWS)
precipitation amount data appears to
present a useful and valid approach to
producing regional analyses of concen-
trations or precipitation amount. The
NWS data were seen to reproduce the
variability in precipitation amount better
than the less spatially dense NADP
precipitation amount data. The precipi-
tation chemistry data appeared adequate
to chart the variation in chemical concen-
trations if the data were first stratified into
fairly homogeneous regions.
Data selection is critical to the results,
more so than previously anticipated.This
process is more critical than general
has been acknowledged because tr
sources of chemical and precipitatic
amount data are numerous. If wi
deposition is the desired regionalize
characteristic, a consideration is whethi
to interpolate the data directly or 1
combine previously interpolated precip
tation concentrations and amount. Th
choice forces an evaluation of tti
representativeness of the various da
sources and the spatial independence
concentration and precipitation amount.
Direct and indirect methods f<
mapping spatial variation wet depositic
were investigated. The indirect methc
allows the use of more representath
precipitation amount data but assume
that precipitation amount and conce
tration are not spatially correlated.
review of experimental evidence su
gests that there is no strong relations)!
between S04 concentration and precif
tation amount across space. It appea
reasonable to interpolate preciprtatii
concentration and convert to local w
deposition fluxes using interpolated ra
gauge data. Further analyses of tl
independence of concentration ai
precipitation amount across space a
recommended, especially where si
specific factors may control wet depo
tion.
Regardless of the choice of data ai
technique, it is important that tl
statistical, meteorological, and chemk
basis for regional analysis be sound.
that regard, it is most likely that infa
disciplinary approaches will produce tl
most useful and accurate region
characterizations of acidic deposits
data.
-------
Richard Vong and Gregory Reams are with Oregon State University, Corvallis, OR
97333, Steven Cline and Joseph Bernert are with NSI, Corvallis, OR 97333,
Donald Charles is with Indiana University, Bloomington, IN 47405, James Gibson,
Timothy Haas, and John Moore are with Colorado State University, Fort Collins,
CO 80523,Rudolf Husar is with Washington University, St. Louis, MO 63130,
Anthony Olsen and Jeanne Simpson are with Battelle Pacific Northwest
Laboratory, Richland, WA 99352, and Steven Seilkop is with Analytical Sciences,
Inc., Research Triangle Park, NC 27709.
Roger Blair is the EPA Project Officer (see below).
The complete report, entitled "Regional Analysis of Wet Deposition for Effects
Research," (Order No. PB 89-181 218/AS; Cost: $13.95, subject to change) will
be available only from:
National Technical Information Service
5285 Port Royal Road
Springfiefd, VA 22161
Telephone: 703-487-4650
The EPA Project Officer can be contacted at:
Environmental Research Laboratory
U.S. Environmental Protection Agency
Corvallis, OR 97333
United States
Environmental Protection
Agency
Center for Environmental Research
Information
Cincinnati OH 45268
Official Business
Penalty for Private Use $300
EPA/600/S3-89/030
0000329 PS
U S SNVIR PROTECTION AGENCY
REGION 5 LIBRAS?*
230 S OEAR80RN STREET
CHICAGO IL
------- |