Project Title: Sentinel Lakes
Project Number: GC11NQ00DRD0000 - 8607DRD - MN 258
Project Chief: Kiesling, Richard L.
Project Personnel: Emily Resseger, Sarah Elliott, Mimi Wallace
Project Start Date: 01-OCT-2008
Project End Date: 30-JUN-2012
MN DNR SLICE (Sustaining Lakes in a Changing Environment)
Project Map
Statement of Problem:
A number of regional and state-wide lake modeling studies have illustrated the potential linkages between climate change, lake morphology, and fish habitat in the form of temperature and dissolved oxygen distributions for Minnesota and the north-central United States (e.g., see summaries in Stefan and others, 1995; De Stasio and others 1996; and in Fang and others, 1999). These models have documented the relative importance of lake-basin geometry, ice-free season, thermal stratification, dissolved oxygen stratification and wind-driven mixing to the development of sustainable fish habitat in deep water lakes of the region. However, the potential trophic-dynamic response to simultaneous changes in climate and land-use is less well understood, as is the response of specific lakes to these historical and hypothetical changes. Questions also remain as to how the complex food webs that support fish guilds within these modeled systems will respond to the predicted physical changes in fish habitat (De Stasio and others 1996).
Calibrated lake models will be used to forecast changes to water quality and deep-water thermal habitat conditions under changing climate and land-use scenarios. Where data are sufficient, models will be used with historic land use and climate data to provide historical benchmarks for comparison with output from scenario models.
Objectives:
- Obtain the physical, chemical, and biological data necessary to develop calibrated bio-physical lake models for three Minnesota sentinel lakes (Carlos L., Douglas Co.; Elk L., Clearwater Co.; and Trout L., Cook Co.) 2.
- Evaluate affects of changing climate and nutrient-loading on lake ecosystem function using bio-physical models.
- Obtain the physical, chemical, and biological data necessary to develop a watershed and lake mixing models to forecast
future water quality conditions in deep lakes with cold-water fish populations.
Relevance and Impact:
Our study is aimed at evaluating the response of sentinel lake indicator species (e.g., Coregonus sp.) to the types of changes in fish habitat predicted by Fang et al. (1999). Fang et al. (1999) used the MINLAKE96 model to simulate suitable fish guild habitat under past (1961-1979) and predicted future (2 X CO2) climate scenarios for 27 Minnesota lake types and three generalized fish guilds. In their study, fish habitat contours were derived from temperature and dissolved oxygen (DO) profiles produced by the MINLAKE96 model output. Contours of fish habitat were compared with temperature and DO habitat criteria derived from the literature for the three broad fish guilds (warm, cool, and cold water fish guilds). Our contribution is to expand on Fang et al. (1999) by using a two-dimensional, biologically and chemically explicit carbon-based model to estimate changes in fish habitat under varying conditions of primary production, allowing us to map deep-water fish habitat based on ecosystem processes.
Strategy and Approach:
Intensive monitoring of three super-sentinel lakes will provide the data necessary to develop predictive lake models in a watershed context. Watershed-loading models will be coupled with in-lake water quality data to calibrate mechanistic lake models. Calibrated lake models will used to forecast changes to water quality and deep-water thermal habitat conditions under changing climate and land-use scenarios. Where data are sufficient, models will be used with historic land use and climate data to provide historical benchmarks for comparison with output from scenario models.
Methodology:
Model data needs will be evaluated using an existing USGS dataset from Shingobee Lake, MN, during the first four months
of the project. Shingobee Lake provides a strong basis for evaluating the capability of the CE-QUAL-W2 to model a
Minnesota Lake with significant groundwater contributions (e.g. Elk Lake) while simultaneously handling the significant
longitudinal transport of water and constituents associated with the Shingobee River (e.g., Lake Carlos). The Shingobee
Lake model will also provide an evaluation of the density of data required to adequately model system behavior.
Data collection for each super sentinel lake will consist of continuous temperature data from multiple depths (e.g.,
thermistor chain with sensors at one-meter intervals, Figure 1) for each major sub-basin in the lake during the ice-free
season as well as one multiple-parameter data platform for the collection of continuous water quality data during the icefree
season. The continuous temperature record from all locations will be coupled with meteorological data collected at the
lake surface to estimate an internal heat budget for the reservoir. Heat flux between discrete depth increments will be used
to estimate vertical mixing of water masses during inflow events. The meteorological data necessary to support this effort
will be collected at a central location in the lake using a moored floating platform (Figure 1). Multiple platforms may be
necessary in Lake Carlos because of its size and complex morphometery.
The data platform will provide the continuous meteorological (MET) and water quality data necessary to populate the
model. MET data collection will include wind speed and direction at the surface of the lake, air temperature, net radiation,
rainfall, and photosynthetically-active radiation (PAR). Water quality data collection from the platform will include standard
four-parameter water quality sonde data (temp., DO, cond., pH) plus chlorophyll a in vivo fluorescence (IVF). Data will be
collected at multiple depths a number of times per day.
Ambient water quality sampling will also take place at regular intervals at up to three sites in the lake and at the inflow and
outflow sites. In addition to the standard sentinel lakes water quality parameters discussed under Result 1, dissolved and
particulate nutrients and carbon as well as algal group abundance will be determined.
Surface water inflow at up to four locations in each watershed will be estimated using temporary, continuous streamflow
gages. Locations will be chosen to provide the best estimates of loading to and discharge from the lakes in question. All
three lakes have well-define outlets. Even-mean constituent loads will be estimated using flow-weighted sampling of
seasonal storm event. Samples will be collected using ISCO automated samplers deployed with independent power
supplies and water level triggers.
Groundwater contributions will be estimated for Elk Lake as part of a separate research study by the University of
Minnesota. Groundwater constituent concentrations will be measured directly using shallow groundwater wells and upgradient
surveys with mini-peizometers. Water balance calculations and model calibration will also provide an estimate of
groundwater contribution in all three lakes.
Progress during FY10
Data was collected from the Super Sentinel Lakes. All three lakes were monitored for influent and effluent nutrient loads, internal nutrient concentrations, lake levels, and physical and chemical field parameters. Inflows, outflows and ambient lake stations at two depths were sampled monthly for nutrients, major ions, chlorophyll-a, and particulate carbon and nitrogen. Continuous lake level data were recorded using pressure transducers deployed at all three lakes. Continuous water temperature data from thermistors deployed at multiple depths were collected in all three lakes. At each monthly sampling, profile data were collected for dissolved oxygen, pH, temperature, chlorophyll-a in vivo fluorescence, and specific conductance. Data from these multiple sampling efforts were coupled with ambient lake monitoring data collected by the Minnesota Pollution Control Agency (MPCA). Together, the combined monitoring data provide 6-8 lake profiles for each lake. Deuterium and Oxygen-18 data were collected from Elk Lake to characterize groundwater inflows to the lake. Sampling included lake surface water, spring inflows from two locations on opposite sides of the lake, and mini-piezometer groundwater samples adjacent to the spring sites at the edge of the lake. Analyses completed include summarizing continuous temperature data to look at the seasonal patterns in thermocline and mixing trends between the lakes. Temperature data were plotted against wind speed to look for relationships between high winds and increased mixing. Basic hypolimnetic oxygen depletion was calculated for each of the three lakes using the oxygen profile data. These calculations provide rates at which oxygen is used in the hypolimnion: an indicator of organic matter accumulation and lake trophic status. The preliminary analyses are intended to assist in developing the biophysical mechanistic models by giving suitable calibration data.
Plans for FY11
Meet with MN DNR project manager to provide a status report, pull thermistor chains for winter, determine feasibility of winter deployment, suervey lake staff gages, continue researching data platform options, meet with MPCA to determine water balance approach using isotope data for Elk Lake, assess results of DNR zebra mussel survey, and begin building lake model.
Statement of Work
Award – July 2009
Data platform deployment in three lakes – August 2009
Inflow water-level gages with water quality samplers – October 2009
Hydrology and water quality data collection – August 2009 through July 2011
W2 model calibration (2009-2010 data) - September 2011
W2 model validation - December 2011
NWIS data publication – April 2012
Final report - June 2012
| 
U.S. Department of the Interior |
U.S. Geological Survey
