|
Home > Research
> NSF
Proposals
Project Description
Results from Prior NSF Support
Comparative Studies of a Suite of Lakes
in Wisconsin
The North Temperate Lakes Long-Term
Ecological Research (NTL LTER) site, established in
1981, is currently in the fourth year of its third grant
period. During the past 14 years we have designed and
implemented a comprehensive study of seven lakes in
the Northern Highland Lake District of Wisconsin and
the surrounding landscape. As evidenced by the 105 peer-reviewed
publications produced in the last five years, we have
made significant advances in the understanding of lakes
and their landscapes (Figure 1). A complete publication list for our project is given in the bibliography.
Some of our most significant accomplishments
in the past five years include:
1. Continuation of the collection and
management of a high-quality, comprehensive, long-term
data set on the physical, chemical, and biological properties
and processes of the 7 lakes and the surrounding landscape
(Magnuson and Bowser 1990).
2. Quantification of groundwater/lake
interactions using stable isotopes and numerical modeling
at individual lake and landscape scales (Anderson and
Cheng 1993, Bowser 1992, Cheng and Anderson 1993, Kenoyer
and Bowser 1992a,b, Krabbenhoft et al. 1990a,b, Krabbenhoft
et al. 1994).
3. Identification of the roles of landscape
position and spatial heterogeneity in explaining interannual
variability of lake dynamics at local (Benson and Magnuson
1992) and landscape (Kratz et al. 1991a, Kratz et al.
in press, Magnuson et al. 1990, 1991) scales.
4. Assessment of how level of taxonomic
aggregation of data influences ability to detect responses
of lake ecosystems to stress (Frost et al. 1992, Carpenter
et al. 1993, Kratz et al. 1994).
5. Development of approaches and protocols
for using remote sensing to map land cover on a statewide
basis for Wisconsin (Lillesand 1993a,b, Bolstad and
Lillesand 1992a,b,c).
6. Use of ice records on lakes as past
and future climatic indicators (Robertson et al. 1992,
Wynne and Lillesand 1992, Wynne and Lillesand 1993).
7. Identification and rough quantification
of the role surface waters play as conduits for terrestrially
fixed carbon to the atmosphere (Cole et al. in prep.,
Kratz and Bowser in prep.).
8. Elucidation of the dynamics of species
richness and assemblage structure in inland lakes (Magnuson
et al. in press, McLain and Magnuson 1988, Tonn et al.
1990).
In addition to direct scientific accomplishments,
our LTER site has been highly successful in catalyzing
interactions with other scientists. The potential for
interacting with LTER researchers and the associated
data bases has been a key factor in generating this
interest. In the past five years alone more than 35
associated research projects, with total funding exceeding
$12M, have been or are being conducted at the site.
These projects are funded from a range of federal (NSF,
EPA, DOE, USGS, USDA-SCS, NASA), state (Wisconsin DNR)
and private (EPRI) sources.
We have also contributed significantly
to education of students. Dozens of undergraduate students
have been involved with research projects at Trout Lake
and Madison. Graduate students have produced 10 MS Theses
and 5 Ph.D. theses related to LTER research in the past
five years.
Our prior research leads directly into
the present proposal both in specific research topics
addressed and in demonstrated abilities to conduct collaborative
intersite research through networking, development of
databases, and conducting workshops for research synthesis
in long-term, regional research. Specific topics on
which this proposal builds are in bullets 3, 6, and
8. Our experience in intersite networking to accomplish
research objectives are in bullets 3 and 7. In particular
we have experience with obtaining the cooperation from
a diversity of research sites for development of intersite
databases on the variation of ecosystems in time and
space. The database, VARiation in North American Ecosystems
- VARNAE, was the first intersite database in the LTER
network and produced at least three peer reviewed publications
(Magnuson et al. 1991, Kratz et al. 1991a, and Kratz
et al. in press).
Understanding Long-Term Temporal
and Spatial Dynamics of Temperate Lakes at Regional
to Global Scales
General rules for ecological systems
must be based upon the diversity of habitats that occur
across broad regional scales. We propose to foster the
recognition of such rules by establishing an informal
global network for examining lake ecosystems across
wide spatial and temporal scales. In our initial work,
we will focus on three questions that can best be addressed
by considering multiple sets of lakes within different
districts. (1) What patterns occur in ecological variability
and organization when lake districts are considered
at regional and global scales? (2) To what extent can
patterns of gains and losses of species from individual
lakes be related to the overall patterns of species
occurrence for lakes within a region? (3) Can analyses
of freeze and thaw phenologies among lakes in the north
temperate zone around the globe be used to describe
and understand dynamics and trends in climate?
The first two questions were selected
to test for organizational properties that operate in
the dynamics of lake ecosystems at a landscape scale.
Each lake district can be considered to provide a replicate
of landscape-level properties and processes. The third
question would take advantage of extensive limnological
records from around the north temperate zone. Answering
it would provide insight into the effects of climate
change and the importance of external forcing factors
in lake dynamics. For our investigations we consider
north temperate lake districts as occurring from about
40o N latitude to the Arctic circle and including
regions with cool to cold winters and mild to hot summers.
We will explore these questions through
a combination of direct contacts and electronic networking.
We will send individual scientists to non-US sites to
work with local data. We will also encourage and facilitate
visits by non-US scientist to our site. Finally we will
hold several international workshops at which long-term
data will be used to address the research questions
above. In all cases we will emphasize collaborations
involving joint analyses of shared data. We will plan
to produce co-authored papers in the peer reviewed literature.
In addition, our efforts will improve the sharing of
ideas and data among long-term researchers on lake systems
around the world.
a. Rationale for Internationalization
The primary reason for seeking the
collaboration of our international colleagues is to
increase both the temporal and spatial extent of analyses
for all of the central questions listed above. For example,
long-term records of North American lake-ice phenology
rarely span much more than a century, while some European
data sets begin in the 1700s and the ice record on Lake
Suwa in Japan begins in 1441 (Lamb 1977). A number of
limnological stations have 20 or more years of comprehensive
limnological data compared with the 13 years of comprehensive
data we have on the LTER project since it began in 1981.
The more extensive records are necessary to detect long-term
trends, identify climatic shifts, and determine stochastic
patterns. Further, a comparison among regionally diverse
sites subject to a range of geological, climatic, and
land-use conditions would provide further insights into
the generality of hypotheses developed at the NTL-LTER
site.
The sites we propose for collaborations
(Table 1) all have a long history
of limnological study. In addition to there being large
variation in geology, land-use and climate among sites,
many have a diversity of lake-types within their region.
For example, the Experimental Lakes Area, situated on
the Canadian Shield, has been a center for experimental
research over the past 25 years. They have consistently
collected detailed limnological information on a variety
of lakes in the region. The region is forested and access
is restricted on many of the lakes. For a detailed description
of the site see Journal of the Fisheries Research Board
of Canada, vol. 28, 1971. The Dorset Research Center
also has a long history of detailed limnological study
on several lakes in South-Central Ontario. The watershed
is continuously forested and human use is primarily
recreational. Lakes are situated on the Canadian Shield
and there has been a history of acid precipitation in
the area (Yan, 1986). Other sites, known for long-term
limnological data, are those we have chosen for collaborative
research outlined in our proposal (Table
1). J. Magnuson, M. Adams and T. Kratz have well
developed relations with the European sites.
Introduction and Objectives:
An early step in a regional analysis
of lakes should be to examine the degree to which lakes
act similarly over different spatial and temporal scales.
Some phenomena will be identified that exhibit consistent
year-to-year patterns across a region. Regional consistency
would identify the spatial scale over which a driving
variable operates. For example, factors controlled by
precipitation would vary consistently across lake districts
that were affected by the same precipitation events.
In contrast, a region wide pattern would not be expected
for driving variables that operate differently from
lake to lake. Such processes would be considered to
function locally. Of course, lakes are affected by many
driving variables some of which act on local scales
while others operate regionally; the behavior of lakes
would be expected to display a complex mixture of these
variability patterns. Additional factors such as the
location of lakes within a district, their connectedness
to adjacent lakes, their size and morphometry, and the
extent of their anthropogenic inputs set a template
that could establish similarity and dissimilarity in
temporal dynamics among adjacent lakes. Analysis of
the degree to which a set of limnological variables
vary over different time scales may also allow inference
regarding the effects of different driving variables.
We propose three themes for analyses.
We will:
1. examine temporal variability patterns
(coherence) of lakes within lake districts following
Magnuson et al (1990),
2. explore the relationship between
landscape position and lake variability across lakes
in different lake districts following Kratz et al.(1991),
3. examine relationships between lake
variability and a variety of other factors such as lake
morphometry and land use.
Comparisons across different lake districts
with long-term data on multiple lakes will provide an
opportunity to determine whether patterns that we have
observed across the Northern Highland Lake District
occur across lake districts worldwide. We hypothesize
that they will.
Background:
Considering temporal patterns, analyses
at the North Temperate Lakes LTER (NTL-LTER) site by
Magnuson et al (1990) demonstrated that adjacent lakes
differed substantially in the interannual variability
patterns exhibited by many variables. For example, some
variables, such as date of ice breakup and lake level,
exhibited consistent variability patterns across lakes.
In contrast, each lake exhibited unique inter-annual
patterns of strong and weak year classes of fish, chlorophyll
concentrations, and water clarity. Adjacency alone did
not lead to interannual coherence, i.e., a consistent
year-to-year pattern of variation across the lake district.
Interannual variability patterns differed systematically
for water chemistry variables. Variability patterns
for nutrients differed substantially among lakes while
elements that were less biologically active or directly
influenced by differences in groundwater input, such
as calcium and potassium, were more consistent among
lakes. In a separate analysis, many variables could
be classified as either lake-specific, those that exhibited
a large degree of consistency within a lake across years,
or year-specific, those that exhibited consistency across
lakes in a particular year (Kratz et al. 1987).
Evaluating spatial patterns, we found
that interannual variability within lakes at the NTL
LTER site was related to a lake's landscape position
for a wide range of limnological variables(Fig 2). For example, the overall interannual variability occurring within
a lake decreased systematically as its elevation decreased
(Kratz et al. 1991). Other patterns related to landscape
position were observed for desert, estuarine and stream
sites, but the pattern was most straightforward for
the lakes.
Chemical characteristics of NTL-LTER
lakes also were related systematically to landscape
position. Hydrologic regimes of lakes there were found
to be influenced by elevation (Cheng 1994, Cheng and
Anderson in press). The proportion of input as groundwater
decreased systematically in lakes with higher elevations
(Swanson et al. 1988) and varying proportions of inputs
as groundwater generated distinct signatures in the
total number of dissolved ions in a lake's water. Analyses
of bog lakes within the region revealed a similar pattern
(Kratz and Medland 1989). In another geographic region,
multi-lake comparisons showed a relationship between
the lakes' watershed areas and their dissolved ion chemistry
(Swanson et al. 1988).
Approach:
We hypothesize that the identification
of temporal and spatial variability patterns across
a wide range of lake districts will lead to a broader
understanding of the general controls over lake processes.
Many of the factors controlling lakes were suggested
by Rawson more than 50 years ago (1939), but our comparisons
of long-term data from sets of adjacent lakes from a
wide range of lake districts will allow us to undersigned
a wide range of controlling factors. By drawing on information
from sets of lakes in different regions we would be
able to test the generality of our observations. With
additional lake data we would also be able to test the
extent that lake size, landscape position, and the degree
of human influence the consistency of variability patterns
exhibited across lakes. We would also be able to explore
how variability patterns differ with the time scales
of observation.
Our approach takes advantage of the
availability of long-term data on sets of lakes within
different lake regions. These data provide the opportunity
to perform analyses not previously possible with long-term
data limited to individual lakes or short-term data
on multiple lakes in a region. Information from a variety
of lake districts could be added to those from our Wisconsin
North Temperate Lakes LTER site. Potential sites include
(1) two Canadian Lake sites in Ontario, The Experimental
Lake Area (ELA) in western Ontario and the Dorset site
in Central Ontario; (2) The English Lake District site
at Windermere run by the Freshwater Biological Association;
(3) Mecklenburg/Brandenburg lake district in northeastern
Germany; (4) the Salzkammergut lake district of Austria;
among others.
Our approach will be to extend our
initial contacts with investigators from the different
set of lake districts to foster a joint analysis of
data sets generated by each site. This analysis will
be catalyzed by an international workshop to be held
at Trout Lake Station located at the NTL-LTER site.
Personal contacts with the two Ontario
sites will be extended through our LTER Augmentation
Proposal and by a number of our investigators. We have
already spoken to the lead investigators at each Canadian
site R. Hecky (Experimental Lake District) and P. Dillon
(Dorset) and they are eager to participate in this program.
For the English Lake District, T. Kratz and J. Magnuson
will extend our interactions. Magnuson spent 4 months
of a sabbatical at the Windermere Laboratory in 1984
and has maintained contacts. Kratz will visit the site,
present a seminar on the topic of the joint research
and work with the investigators interested in participating
with us in this analysis. M. Adams has already developed
interactions with the German and Austrian sites through
a funded international proposal on primary production.
We also plan to hold a general comparative
limnology workshop that would bring together investigators
who have collected information from sets of lakes within
regions. We plan to explore the general utility of employing
a landscape-scale limnological perspective. This workshop
would be held early in year 3.
Introduction and Objectives:
Factors that regulate biodiversity
and the distribution of species in time and space are
important because increasing human populations and their
increased resource use are stressing the worlds biota.
This is evidenced, for example, by an increasing number
of species being added to endangered species lists.
Human activity causing alteration and loss of natural
habitat are unrefuted sources of species extinctions
and invasions. In addition, there is a background level
of species turnover driven by non-anthropogenic forces.
It is important to understand these natural dynamics
in order to evaluate the extent of human effects on
biodiversity. For lakes, research on processes controlling
diversity has largely been focused on individual systems.
Emphasis has focused on how competition, predation,
and physiological limitations imposed by the environment
influence productivity, succession and other in-lake
processes. We propose a broader approach involving the
investigation of gains and losses of species and their
turnover rates on expanded spatial and temporal scales.
We propose to interact with a group of scientists in
North America, Europe, and the People's Republic of
China to assemble long-term data from a variety of lake
districts and accomplish the following objectives:
1. to investigate regional dynamics
of species distribution as related to species gains,
losses, and turnover rates in individual lakes in different
lake districts,
2. to relate species dynamics to the
extent and strength of interconnections among lakes
within different regions,
3. to investigate how human activities
may influence these processes.
We will determine the extent to which
a region's species distributions and turnover rates
can be related to its geomorphic and geochemical characteristics.
We will also evaluate whether trends in species gains,
losses, and turnover rates are consistent when analyses
are extended over longer time scales.
The central objective of our investigations
will be to evaluate patterns of regional dynamics across
to lake districts around the world. We will examine
the extent to which regional dynamics are related to
the different geomorphological and geochemical characteristics
of the lake districts. We will test the generality of
our observation at the NTL-LTER site that regional patterns
of species diversity and distributions tend to remain
constant over time despite relatively unstable local
dynamics (Cisneros 1993, Magnuson et al. 1994, S. Arnott
unpublished data).
By comparing patterns of species gains
and losses across different lake districts, we will
also be able to evaluate how differences in land-use
patterns and climate influence regional species dynamics.
In addition, using information from long-term data bases
we will determine how consistent trends in species gains
and losses are over time. This will help provide insights
into the 'invisible present' (Magnuson 1990) and reveal
how comparable short-term dynamics are to long-term
patterns.
Background:
Metapopulation models have provided
a useful framework for investigating the regional distribution
of organisms with unstable local population dynamics
(Hanski et al. 1994, Collins and Glenn 1991). A strength
of the metapopulation approach is that it emphasizes
the dynamic and interactive behavior of populations
over time. Despite unstable population dynamics (frequent
extinctions and re-introductions) within individual
habitat patches, populations can persist on a regional
scale through the dispersal of organisms within the
metapopulation (Hanski 1989, Harrison and Quinn 1989,
Quinn and Hastings 1987). Thus, despite high local extinctions
through time, the probability of persistence of a species
on a broader scale is increased by high dispersal rates
among habitat patches. Present studies at the NTL-LTER
site indicate that the species composition of zooplankton
and fish assemblages is dynamic, with gains and losses
of species occurring annually (S. Arnott, Magnuson,
et al, in press). Metapopulation theory may provide
useful insights into these local fluctuations in species
presence and absence and their role in the regional
dynamics of species distribution.
Changes in the regional distribution
of species are dependent on extinction and immigration
of species in individual habitats. We can predict these
changes in distribution using a generalized metapopulation
model (Gotelli and Kelley, 1993),
dp/dt = g(p)*(1-p) - h(p)*p
where p is the proportion of sites
occupied and g(p) and h(p) are functions describing
the probability per unit time of immigration and extinction,
respectively. These functions can have several forms;
1) constant, indicating that immigration and extinction
rates are constant for each site; 2) dependent on the
proportion of sites occupied, indicating that regional
occurrence of species influences the species turnover
rates; or 3) complex, indicating non-linearities in
the relationships between spatial variation and the
size of habitat patches (Hanski and Gyllenberg 1993).
Fitting these models to empirical data may provide useful
insights into dispersal patterns of taxa in the various
regions to be studied.
Metapopulation models have the potential
to incorporate the importance of interaction among lakes
within a region. Few studies have considered biodiversity
in lake districts from a metapopulation perspective.
Determining patterns of species disappearances and appearances
will enable us to evaluate the processes underlying
regional species distributions. We will be able to evaluate
relationships between the extent of connection among
lakes and overall patterns of regional dynamics.
Approach:
One of our earliest tasks will be to
formalize our interactions to foster collaborations
and data exchange among identified sites with long-term
data sets of lakes within a region. We have already
contacted many sites interested in participating but
we will continue to expand that list as other sites
are identified (see Table 1 for list of potential participating
sites). At many of the participating sites, Shelley
Arnott, a graduate student at the NTL-LTER site, will
present a seminar providing an overview of research
at the NTL-LTER site and the questions we hope to address.
She will interact with interested participants, identify
appropriate data and arrange for database transfers.
Contacts have already been extensive with the Dorset
Research Centre in Ontario. Dr. Norman Yan has expressed
his interest and has already transferred his extensive
zooplankton database to Madison. Analyses using the
Dorset zooplankton data are presently underway.
A team of American investigators, including
J. Magnuson plan to visit the Lake Balaton site in Hungary
in May 1994, with a follow up scheduled for fall 1994.
At that time, participants will be identified and arrangements
for data transfer negotiated. Dr. Xie Ping, a zooplankton
ecologist from the Institute of Hydrobiology, Chinese
Academy of Science will be a Visiting Scholar the Center
for Limnology for 9 months beginning summer 1994. At
that time, we will discuss his participation in the
project and make arrangements for data acquisition.
Other sites, including the Experimental Lakes Area,
the English Lake District and the Limnologisches Institut
will be visited spring 1995.
Our analysis will focus primarily on
zooplankton and fish species. Annual gains and losses
of aquatic species in individual lakes will be investigated
to determine the nature of change in species composition
over time. These calculated gains and losses will be
used to investigate the dynamics associated with the
regional distribution and diversity of species (e.g.
Fig. 3). Rates of species turnovers will be correlated
with factors such as distance to other lakes, lake size,
geomorphic characteristics of surrounding areas, and
human activity on lakes within each region. Results
will be compared among regions to elucidate commonalities.
There are a variety of things to consider
when evaluating gains and losses of species from zooplankton
and fish data. For our purposes, we will consider a
loss to have occurred if we do not detect a particular
species in the water column after having observed it
in samples from the previous year. For example, a zooplankter
that disappears from the water column will be considered
locally extirpated even though resting eggs or encysted
zooplankton may have been deposited in the sediments.
Another important consideration is that rare species
may be underrepresented in samples. Rarity occurs both
temporally and numerically and can be dealt with, to
a certain extent, by increasing sampling effort (Green
and Young 1993). Rarity associated with temporal fluctuations
in abundance may be alleviated by decreasing intervals
between sample dates thereby increasing the probability
of sampling on date when a particular species was abundant.
Data used for these analysis will integrate information
taken from frequent intra-annual samples, thereby increasing
the probability of detecting rare species. We will analyze
how the detection of rare species influences species
immigration and extinction dynamics by examining the
relative contribution of low density species to the
overall dynamics.
d. Lake Ice
Phenology, Climate Change and Variability
Introduction and Objectives:
The rapid increase in atmospheric
concentrations of greenhouse gases since the industrial
revolution has been well documented and it is generally
believed that this increase will cause substantial changes
in global climate. For example, most general circulation
models (GCMs) predict that global average temperatures
will increase by 1.9 to 5.2[[ring]]C from pre industrial
conditions, given a concomitant doubling of atmospheric
carbon dioxide from pre industrial levels (National
Academy of Sciences 1991).
Detecting actual greenhouse warming
effects has been difficult. Researchers have attempted
to detect these effects using two different approaches:
1) analysis of surface and satellite based temperature
measurements, and 2) assessment of a variety of indirect
climate indicators. Both GCMs (Manabe and Wetherald
1986, National Research Council 1982) and observations
(Groisman et al. 1994) predict that the magnitude
of temperature increases due to enhanced greenhouse
warming will be greatest at higher latitudes and that
winter and early spring temperatures will manifest a
relatively greater increase than summer temperatures
in those latitudes.
Lake ice cover is likely to be well
suited as an indicator of such changes because of:
1) Its intrinsic link with winter at
higher latitudes,
2) Its proven utility as a climate
indicator (e.g., Palecki and Barry 1986, Robertson et
al. 1992, Schindler et al. 1990, Wynne and
Lillesand 1993),
3) The suggested trends toward decreased
ice durations (primarily due to earlier breakup dates)
on several inland lakes and Great Lakes bays (Comb 1990,
Schindler et al. 1990, Etkin 1991, Hanson
et al. 1992, Robertson et al. 1992, Robertson
and Assel 1993),
4) The availability of long- term data
sets pre- dating temperature measurements in some areas,
5) Its ability to be detected from
the space borne sensors from which substantial archival
data are available (e.g., Maslanik and Barry 1987, Wynne
and Lillesand 1993),
6) The availability of excellent spatial
coverage, in terms of both distribution and density,
in comparison with either other climate proxies (e.g.,
glaciers) or direct temperature measurements,
7) Its utility as a climate proxy in
the high latitude areas of the Northern Hemisphere which
have a relative paucity of meteorological stations (Palecki
and Barry 1986, Robertson et al. 1992, Wynne
and Lillesand 1993).
Our general objective is: To analyze
interannual variability in the spatiotemporal distribution
of lake ice phenology in the Northern Hemisphere to
describe and quantify, where possible, the antecedent
climatic variability.
Our specific objectives are to determine
whether patterns in lake freeze and thaw dates observed
in Wisconsin are general for larger regions of North
America and the northern hemisphere. Particular patterns
are the recent warming observed in the thaw dates and
the pattern associated with ENSO events (Robertson 1989,
Anderson et al. in progress), interdecadal changes in
the strength of the Aleutian low (UNESCO 1992), and
the inverse relation between winter severity between
Asian and England (Lamb 1977). Other expected . Other
expected patterns will be developed from the modeling
efforts of J. Foley for patterns in global climate both
around the northern hemisphere and over time.
Background
Ice records for Lake Mendota (Robertson
1989b, Robertson et al 1992) in Madison, Wisconsin is
one of the longest continuous record (135 years) in
North America and it pre- dates the temperature record
for the area. Analysis of this record (Fig. 4) suggested a strong correlation between the duration
of lake ice cover and the
end of the Little Ice Age at the end
of the nineteenth century. More importantly, Robertson
detected evidence of a general warming trend since the
beginning of this century (Robertson et al. 1992).
The recent warming pattern as well as the presence of
an el Niño signal seem to be general for a larger
set of Wisconsin Lakes in our current efforts (W. Anderson
et al in preparation) .
Wynne and Lillesand (1993) have assessed
the capacity of space borne sensors to detect ice in
small inland (Wisconsin) lakes. Substantial archived
data are available from these sensors. We used data
from the Advanced Very High Resolution Radiometer to
discriminate the presence and extent of lake ice during
the winter of 1990-1991 on the 45 lakes and reservoirs
in Wisconsin with a surface area greater than 1,000
hectares. Our results indicate the feasibility of using
the AVHRR to determine the date of lake ice breakup.
Also, a strong correlation (R = -0.87) exists
between ice-off dates and local surface- based temperature
measurements.
Wynne and Lillesand (1993) are in the
first year of a three- year project, funded by the Great
Plains Center of the National Institute for Global Environmental
Change (NIGEC), that focuses on satellite monitoring
of lake ice as a climate indicator in the Western Great
Lakes region. The specific objectives of this project
are to 1) extend the temporal and spatial range of earlier
results by compiling relevant satellite image data for
the region, 2) undertake spatially explicit retrospective
(process) modeling and trend analysis of lake ice phenology
throughout the study area, and 3) to perform modeling
to simulate lake ice phenology under alternative future
climate scenarios throughout the study area. The process
modeling component of the NIGEC project provides a basis
for attributing any changes in ice phenology to attendant
climatic causes. S. Hostetler (USGS) has written a physically
based eddy diffusion model to simulate the formation
of lake temperature and evaporation. This model includes
a thermodynamic submodel to simulate the formation of
lake ice and the melting of ice and snow and can be
used with climate models to develop scenarios of expected
changes in lake ice under various climate scenarios.
The internationalization of the ice
phenology research will greatly enhance our ongoing
research activities. Currently, there are several projects
at the UW-Madison that examine relationships between
climate and the thermodynamics of lake ice. These include:
(1) the aforementioned climate / lake ice phenology
study of Wisconsin (Robertson, 1992; Wynne and Lillesand,
1993), (2) development of
globalclimate and hydrological process
models (Foley) and (3) examination of the role of lakes
in influencing climate using GCMs (in collaboration
with Dr. Gordon Bonan at the National Center for Atmospheric
Research).
Approach:
We intend to take advantage of the
long-term ice freeze and thaw data available in the
northern hemisphere through international networking
and synthesis workshops. These data, based on our experience
in Wisconsin are lying fallow in the files of many agencies
and research laboratories across the northern hemisphere.
Our first task will be to renew or establish the contacts
required for collaboration and data exchange among the
identified sites (Table 1) with long-term data on lake-ice
phenology. J. Magnuson will visit several sites in fall/winter
1994 to present a seminar giving an overview of the
related NTL-LTER research and the specific goals of
this study. During that time sources of long-term lake
ice records will be identified and some workshop participants
will be identified. Additional ice freeze and thaw data
will be identified through contacts made at the Society
for International Limnology meeting in summer 1995 and
sources listed in Lamb (1977) for Japan and Sweden.
The analysis and synthesis workshop will be held at
the NTL- LTER field site (the University of Wisconsin-
Madison Trout Lake Station) in 1995. The workshop will
be one week in duration and will be focused on analyzing
and synthesizing information on ice durations for the
northern hemisphere. Workshop participants will be encouraged
to submit their ice phenology data sets prior to the
actual workshop so that the necessary data entry and
pre-processing will be completed prior to their arrival.
This will also afford us the chance to analyze the aggregate
dates with both traditional time-series as well as process
modeling approaches prior to their arrival.
We propose to examine the relationships
between the spatiotemporal variability in climate and
lake ice phenology by using two techniques: (1) statistical
analysis of historical ice-phenology data that span
the circumpolar temperate region and (2) process-based
modeling of lake-ice responses to climatic forcing.
In the first technique, we will use a variety of statistical
methods to compare the historical ice data to known
modes of climatic variability, including the El Niño
/ Southern Oscillation (ENSO), interannual oscillations
in the Aleutian low (Trenberth 1990) and interdecadal
variations in the coupled atmosphere-ocean system (Schlesinger
and Ramankutty 1994). In this work, we will apply singular
spectrum analysis (Broomhead and King 1986, Fraedrich
1986) and more traditional statistical techniques (Robertson
et al. 1992). We expect that the historical lake-ice
data will show complex responses, in both space and
time, to climatic variability.
To better understand these historical
data, we will apply process-based models including the
lake thermodynamics model of Hostetler (1991), to a
spatial grid that spans the circumpolar temperate zone
(from 30 to 90[[ring]]N). For all models we will use
several different estimates of historic climatic conditions
(i.e., long-term means, composites of ENSO years, and
empirical orthogonal functions (EOFs) of inferred modes
of climatic variability). Thus, we hope to better explain
the observed patterns in the historical lake ice data.
We also will apply the lake thermodynamics models to
GCM based projections of climatic change.
Professor Jonathan Foley (University
of Wisconsin- Madison Department of Atmospheric and
Oceanic Sciences and Director of the Institute for Environmental
Studies Climate, People, and Environment Program) has
agreed to collaborate on this international synthesis
activity, including helping to supervise the graduate
student that we are proposing to pay from both this
proposal (0.25%) and from existing funds (0.25%). This
student will be primarily responsible for the re-examination
of our workshop results using a lake-ice/general circulation
model.
|