Loading...
HomeMy WebLinkAboutLake Cornelia Geol student paperLAKE CORNELIA GEOCHEMICAL TRACING OF LAKE HEALTH 20 Dec 2018 Anita Hall, Sarah Howe Lake Cornelia, located in the heart of a residential and commercial community, is a water body characterized by high nutrient loading, poor water clarity, and periodic toxic algae blooms. These conditions have potential to impact public health as well as restrict recreational use of the lake. The purpose of this project was to determine how environmental conditions and water quality in Lake Cornelia have changed over time in order to provide context for future rehabilitation efforts. Carbon and nitrogen analysis, x-ray fractionation, and diatom analysis were performed on sediment cores from the north basin of the lake. Carbon and nitrogen data, as well as the diatom analysis, suggest that the basin was a much drier wetland or peatland environment for a period of time, but returned to a lake state midway up the core. Trends in copper, zinc, lead and arsenic indicated probable spikes in human activities over the course of the last century. Our historical research helped us draw the conclusion that rapid urban expansion from Southdale mall and residential area in the 1950s and 60s probably introduced levels of these elements to the surrounding environment and that large-scale manipulation of the basin’s watershed by the installation of storm sewers during this time was likely partially responsible for the basin’s return to a lake state. Humans have drastically altered the geochemistry and watershed of Lake Cornelia, making the level of rehabilitation that is achievable largely dependent on the management of contaminant and nutrient influxes from the surrounding urban community. Lake Cornelia Page 1 INTRODUCTION Lake Cornelia is a small, 58-acre shallow lake located in the southeastern quadrant of Edina, Minnesota, within the Nine Mile Creek Watershed District (NMCWD). (Figure 1) At the turn of the twentieth century the land around the lake was “far too soggy for much more than cattle and sheep grazing...” (Morse-Kahn, 2009) The nearby spring that fed Nine Mile Creek provided cold spring water to the surrounding farms and marsh. By 1950 the land was about to be transformed. Dayton Development Company purchased 500 acres, encompassing Lake Cornelia and extending west to today’s York Avenue. Plans were unveiled in 1952 for the creation of the first enclosed, climate- controlled shopping mall with an adjacent new housing development complete with a park and land dedicated to other commercial ventures. (SLP Historical Society and Hesterman, 1988) In 1956, Southdale Mall opened along with newly constructed County Road 62 (now Crosstown 62) to manage traffic flow. Within 10 years, the watershed surrounding Lake Cornelia had been dramatically and permanently altered. Today Lake Cornelia shows the stress of urbanization. What was once sited as a “clear body of water without marshland” (Morse-Kahn, 2009) is now a hypereutrophic lake which is on the EPA’s list of Impaired Waters. (MPCA, 2018) The lake was treated in 2017and 2018 for a Curly Leaf Pondweed invasion and in 2016 and 2017 was temporarily closed for a high level of blue-green algae, which released the toxin microcystin. (Walsh, 2017) The lake has been monitored over the past decade by the Metropolitan Council Citizen-Assisted Monitoring Program (CAMP) and has been found to contain high levels of both total phosphorous and chlorophyll a as well as poor Secchi disc transparencies (153 μg/L, 51 μg/L and 0.4 m). According to the NMCWD’s Water Quality Classification, these measures have placed the lake in a Level IV rating, which is Generally Intended for Runoff Management. The goal for Lake Cornelia is to improve the quality measures to attain a Level III rating, which is defined as Supports Fishing, Aesthetic Viewing Activities and Wildlife Observation. (Lake Cornelia UAA, 2010). Barr Engineering conducted an extensive analysis of Lake Cornelia in 2004 and 2008 for a Use Attainability Analysis. (Lake Cornelia UAA, 2010) Along with NMCWD, Barr issued five goals for the lake. In addition to improving the lake to a Level III rating, the lake also needs to provide runoff capacity for a 100-year flood event, achieve water quality that will support a ‘diverse and balanced native ecosystem’, provide for recreational usage and protect existing wildlife. The report focused on measures that could be taken to attain the Level III rating. This would require bringing the total phosphate to < 60 μg/L, which would in turn reduce the algae in the lake and improve clarity of the water. Barr outlined twelve Watershed Management Best Practices which would improve the lake water quality. A recommendation for future work was to extract a deep sediment core from the lake in order to evaluate how past changes to the land have altered the lake’s water quality. This will enable NMCWD to determine if it is possible to return the lake to a higher standard of water quality. The objective of this research therefore is to extract both a short and long sediment core from north basin of Lake Cornelia and examine them to determine how the lake has changed with time, specifically looking for changes to organic plant matter and to contaminants. We will attempt to correlate these changes to the modifications of the land use around Lake Cornelia and answer the following questions: How has the dominant vegetation in the lake changed over the past century? How has the geochemical signature of Lake Cornelia changed over the last century and can these changes be correlated to land use changes and/or man-made inputs? Do changes in the diatom population confirm changes in the lake’s history? Lake Cornelia Page 2 MATERIALS AND METHODS Sediment Cores: Sampling was conducted on the north basin of Lake Cornelia during Oct 31 and Nov 2. 2018. Using a Glew Gravity Corer, a short sediment core was drawn from water that was about 1.2 m in depth. (Figure 2) The water-sediment interface was preserved, and 24 cm of sediment were captured within a 6 cm diameter Lexan tube. The core was maintained at ambient temperature in a vertical position until Nov 6 when it was sectioned into 1 cm cylinders which were then sampled and refrigerated. A Square Rod Livingston Piston Corer was used to obtain a deeper sedimentary core taken from roughly the same area. On the first drive, hard packed sediment was encountered which we were unable to penetrate. A 60 cm long core of 6 cm diameter was obtained. The core was capped and maintained in a horizontal position for approximately 3 hours until it could be extruded into a PVC sleeve and stored at 4.4°C. The core was split, described, and sampled at 1 cm intervals. (Figure 3) Sediment samples collected from both cores were dried for a minimum of 48 hours at 100°C in a Fisher Scientific Isto-Temp oven, then ground to a fine powder with a ceramic mortar and pestle. Carbon and Nitrogen Analysis: A Flash EA 1112 NC Soil Analyzer, with combustion temperature of 900°C, was used to determine the carbon and nitrogen composition of each sample. 3-4 mg from each of the dried and ground sediment samples were placed inside a 5x8mm silver foil cup. Each cup was treated twice with 6.4% sulfurous acid, (30μl followed by 50μl) in order to react and remove any inorganic carbonates. No notable reaction occurred in any of the samples. Between each acid treatment, the samples were dried at 100°C in the Isto-Temp oven then crimped closed. The samples were analyzed in four batches each consisting of twenty-seven samples, three aspartic acid standards and two blanks. X-ray Fluorescence Analysis: A hand-held InnoveX XRF Analyzer was used to determine the percentage of various elements in our samples. The XRF was set in a fixed stand within a lab setting. Soil Assay mode was used for analysis. Dried and powdered samples from the 1 cm sections of each core were too small for the XRF analyzer to process, so samples from two adjacent sections of each core were combined then placed into a TremLess ® sample cup. Prior to the run the unit was calibrated with a calibration disk. All standard test protocols were followed. Slides: Small samples were extracted from locations in each visibly differentiable (by color, texture, density etc.) section of the core at 36, 47, 58, 78cm. Smear slides were made from each of these samples and examined under a microscope at 40x. Lake Cornelia Page 3 RESULTS An image of the Livingston core as well as a full description of its layers are included in Figure 3. Full length core profiles (Figures 4, 5a & 5b) were constructed by stitching together the Glew and Livingston core data. Overlap was estimated based on correlation in carbon (C), nitrogen (N) and XRF data. Carbon and Nitrogen Analysis: Downcore from 48 cm the C and N values are relatively high, except for a dip near 57 cm. C and N drop noticeably upcore from 48 cm and remain low until 5 cm, where they gradually increase to present levels (Figure 4). The C:N ratio displayed no coherent trend over the course of the core, fluctuating within a 5-unit range between 10 and 15. The spike in C and N values at the base of the Glew core could be due to contamination during the coring and/or extrusion processes. The average deviation from the standard for N was 0.4974% with a standard deviation of 1.3086, and for C the average deviation was 0.8439% with a standard deviation of 3.2277. The average deviation from the blank for N was 0.000%, and for C the average deviation was 2.496% with a standard deviation of 5.527.1 Prior to the first run, the analyzer calibrated with aspartic acid. The resulting calibration curve yielded an R2 of .999738 for C and .999564 for N. X-ray Diffraction Analysis: Average error was 4.5%.2 Figure 5a displays the results for copper (Cu), zinc (Zn), lead (Pb), and arsenic (As) compared to the level 1 and level 2 Sediment Quality Targets (SQT) outlined in Table 1 and detailed in Crane and Hennes (2016). The four listed elements each currently exceed the level 1 target. The concentrations of both Cu and Zn in the uppermost surface sediment are approaching the level 2 SQT and their concentrations are increasing with time. Both Pb and As reached their peak at 40± cm downcore and are now below the level 2 target. At 50 cm downcore, the Cu concentration in Lake Cornelia’s sediment began to climb from a background level of ~25 ppm to a high of 111 ppm at the top of the core. The rate of change was 0.31 ppm/cm. Zn follows a similar pattern to Cu. At 50 cm downcore, the Zn concentration in Lake Cornelia’s sediment begins to rise from a background level of ~90 ppm to a high of 424 ppm at the top of the core at a rate of 10.5 ppm/cm. The surface concentration is approaching the level 2 SQT. Also, at 50 cm downcore, the Pb concentration in Lake Cornelia’s sediment suddenly rises from a background level of ~11 ppm to a high of 198 ppm where it remained constant for 6 cm. The top 25 cm of the core show a slow decline in Pb concentration to the present value of 123 ppm. It is only in the top few cm of the core that Pb concentrations once again fall below the level 2 SQT. At 55 cm downcore, the As concentration in Lake Cornelia’s sediment begins to rise from a background level of 6 ppm to a high of 49 ppm reached at 44 cm downcore. The As concentration remains at a high level for about 5 cm, then starts to slowly decline to the present value of ~18 ppm. Figure 5 b displays the XRD results for calcium (Ca), titanium (Ti), chromium (Cr) and iron (Fe). These four elements are common products of weathering of rock. Ti, Cr and Fe each show a jump in concentration at 68 cm downcore, followed by a drop at 60 cm. These three elements follow a similar trend throughout the core, including a rise in concentration with decreasing depth throughout the Glew core. Ca follows an inverse relationship from Ti, Cr and Fe, including a decrease in concentration with decreasing depth of the Glew core. Diatom Analysis 5 smear slides from different locations in the Livingston core provided a look at what types of organisms existed at select moments in time. (Figure 6) The smear slide taken at 36 cm from the fine-grained lake sediments was populated with planktonic diatom species. One species of diatom found at 36 cm strongly resembled Nivicula trivalis, a species known to increase as nutrient levels rise in lakes in Maine (Danielson, 2009). Smear slides below the transition to denser, vegetation rich soils (47, 58, and 78 cm) showed a lack of planktonic diatoms and were instead dominated by sponges and plant material. 1 High average and standard deviations are likely due to the inexperience and unfamiliarity of the author with data collection. 2 Calculated error for XRF measurements for the sediment and deep core are given in parentheses for each element: S (7.3%, 9.0%); Cl (18.7%, 11.5%); K (1.5%, 1.6%); Ca (0.9%, 0.9%); Ti (1.4%, 1.3%); Cr (6.8%, 7.6%); Cr (6.7%, 7.6%); Fe (0.3%, 0.3%); Cu (3.1%, 6.5%); Zn (1.2%, 1.1%); Pb (1.2%, 2.3%); As (7.7%,6.6%) Lake Cornelia Page 4 DISCUSSION Interpreting the data obtained from the core requires an attempt at dating the core. Without the financial means to conduct radiometric dating or pollen analysis, we turned to historical events and attempted to find correlation. The great drought of 1929-1941, displayed in MN PDSI data in Figure 7, may have created conditions which lead to the lake’s disappearance. This dry period may account for the dry peat-like lower section of the Livingston core and the fact that we were unable to retrieve more than 60 cm of sediment with the Livingston corer. This change from a water filled lake to a dry peat basin is supported by the CN data (Figure 4). The shift at 48 cm from higher levels of C and N to lower levels upcore could indicate a shift back to a wetter time with the lake reappearing. There is another notable dip that occurs on the C and N data at 58 cm. This could correspond with the peak of the great drought, but no concrete conclusions can be drawn. We were not able to see significant changes in vegetation over time using C:N ratio, the values were in the range of aquatic plants over the entire span of the core. The opening of Southdale Shopping Center and the nearby housing development in the mid-1950s supports the hypothesis that 45-50 cm downcore represents the 1950s timeframe. (Figure 9) If this is true, it might help explain the sharp transition from a coarse, peat-like texture to fine wet sediment. Lake Cornelia accepts storm sewer discharge from Southdale Shopping Center and the surrounding residential areas. (Figure 10 and Lake Cornelia UAA, 2010) The transition from soil to fine sediment could indicate when storm sewers began to discharge into Lake Cornelia in the 50s and 60s. The increased drainage into the lake may have played a part in the lake’s transition from a wetland or peatland (which would produce drier, vegetation rich soil) to a lake (which would produce fine sediment). Our inferences about the basin’s dry and wet periods were further supported by our diatom analysis. Planktonic diatoms in the fine sediment segment of the core indicate that an aqueous environment (in this case, lake) has been the dominant state in recent years. Below 48 cm, where the core shifts to drier, spongy, and vegetation rich soil sediments, no planktonic diatoms were found. This supports the conclusion that below this transition, Cornelia was not a lake environment. Diatoms found in smear slides from the recent lake-state period include a species that strongly resembles Navicula trivalis (Figure 6, Image H). Navicula trivalis is a species of diatom that was found to increase with nutrient loading in lakes in Maine. (Danielson, 2009) This species could be related to the diatoms we found, which would make sense considering the high nutrient concentrations in Lake Cornelia. XRF data gave us insight into human activities’ impacts on the water quality in Lake Cornelia (Figures 5a, b). The metals Cu, Zn, Pb and As show a notable increase in concentration starting at ~50 cm, about the same location where the transition to lake sediments occur. This prevalence could indicate development around the Lake Cornelia drainage area, as these elements often indicate increased human activity. A possible source of Cu is brake pads. Brake pads were determined to contribute 39 ± 21% of the Cu to the watersheds in San Francisco.3 Other sources of Cu could be algaecides, fertilizers, pesticides, pressure treated lumber, architectural Cu, and Cu weathered from rocks. (Crane and Hennes, 2016) Major sources of Zn in industrial runoff are galvanized metal (storm sewer pipes, ducts, light poles, gutters), motor oil leaked onto pavement, tire dust (tires contain ~1% Zn by weight). Other sources of Zn are white paints (zinc oxide) and wood preservatives. (Golding, 2008) Tetraethyl lead was added to gasoline starting in 1922 to increase fuel economy and reduce wear on engine parts. With the passage of the Clean Air Act in 1970, Pb levels in gasoline were reduced. Cars that utilized catalytic converters, requiring unleaded gas, were introduced in 1975, thus beginning the phase out of leaded gas. In the US, there was a total ban on the sale of leaded gas starting in 1995 (Fowler, 2008). These regulations most likely resulted in the gradual decline in Pb levels above 40 cm in the core. Possible sources of As include sodium arsenite, which was used in MN in the 1950s and 1960s to control aquatic vegetation and agricultural pests. (Crane and Hennes, 2016) Attention to As as a toxic pollutant has resulted in regulating its use, which may explain the decline in As levels in the core above 40 cm. The spike in metal concentrations occurred at about the same location in the core as 3 San Francisco has had difficulty meeting the Clean Water Act requirement for copper in urban runoff waters. In 2010 California and Washington passed laws limiting the amount of copper that can used in automotive brakes to no more than 0.5%. (Donigan and Bicknell, 2007) In response the Brake Pad Partnership recommends that copper use in brake pads be phased out by 2033. (CDA webpage) Lake Cornelia Page 5 the transition to lake sediments. For this reason, is important to note that the concentration of metals in the soil may have also been impacted by the installation of storm sewers which may have provided a mechanism by which metals were transported from the surrounding environment into Lake Cornelia’s basin. We chose not to focus on interpreting causes for the trends in elements calcium, titanium, chromium, and iron. However, their synchronous fluctuations deep in the core could indicate response to changing redox conditions in a wetland environment. Through outreach to former Edina employees, we discovered that the Cornelia basin was briefly farmed and grazed. This is a major cause of disturbance which should be kept in mind when viewing our cores. It is likely that there is an unconformity in the core at this time period due to the soil removal and tillage. Far away sources of soil and dust also could have been added during the “dust bowl” (Steil, 2016 interview with John Handeen)4. 4 “…Handeen can point to changes in the land from the dust storms … where blowing dirt accumulated like snow drifts in winter …” (Steil, 2016 interview with 91 yr. old Montevideo native, John Handeen) Lake Cornelia Page 6 CONCLUSION How has the dominant vegetation in the lake changed over the past century? Do changes in the diatom population confirm changes in the lake’s history?  We were not able to make meaningful conclusions about changes in vegetation over time using C:N ratio. However, the C and N values began a rapid decline at the core’s transition from peat to lake sediments, supporting the conclusion that the basin was once a wetland and began to transition back to a shallow lake at this point in time. Our diatom analysis further supported this conclusion.  Lake Cornelia has been drastically altered from its natural state. Historical aerial photos, historical documents and the memories an older Edina resident5 give the best indication that the lake has undergone three major transitions. In the mid-1800s the lake was part of an extensive natural wetland. As settlers arrived and began to farm the area, the wetlands were drained and sometime during the 1930-1940s the lake dried up and was used for grazing. The lake acquired its current state in the 1950s with the buildup of Southdale and the surrounding residential area. During this period of urbanization storm sewers were added, altering the watershed’s drainage and making storm runoff a major component of the lake’s inflow. How has the geochemical signature of Lake Cornelia changed over the last century and can these changes be correlated to land use changes and/or man-made inputs?  Storm drainage, watershed manipulation and urban growth have negatively impacted Lake Cornelia. When storm sewers were added, the natural flow of water into the lake was permanently altered. (Figure 11) With the run off, came urban pollutants such as Cu, Zn, Pb and As.  Cu and Zn pollution should be studied further as they are approaching level 2 SQT. Cu and Zn can accumulate in fish tissues, interfere with reproduction, weaken the immune system, and “induce pathological changes.” These heavy metals and metalloids can be harmful at low concentrations and will bioaccumulate in the food chain. (Crane and Hennes, 2016)  Considering our data, it is our conclusion that unless changes are made to the storm drainage system into Lake Cornelia, rehabilitation efforts will be limited as it is essentially a drainage pond for an urban area. Our data provide historical context for future rehabilitation efforts in Lake Cornelia. We would like to acknowledge and thank the following individuals for their help in completing this project: Bob Kojetin, Edina’s former Parks and Recreation Director, for answering our questions about what Lake Cornelia looked like in the 1940s; Gael Zembal, the Education and Outreach coordinator for the Nine Mile Creek Watershed District, who suggested that we study Lake Cornelia and who provided us with key documents; Ryne Regger, a senior from Edina High School, who was our field assistant and Kevin Theissen, our professor who helped us extract the Livingston core while floating on a sinking raft without getting any of us wet, and who also patiently answered numerous questions. 5 Conversation with Bob Kojetin on 25 Nov 2018: Bob is 85, a longtime resident of Edina and the former Parks and Recreation Director for Edina. Bob confirmed that when he was a kid the area that is now Lake Cornelia use to be farmed and there was a 3’ wooden culvert that enabled cows to walk beneath 66th street. Lake Cornelia Page 7 FIGURES AND TABLES Figure 1 – Lake Cornelia, Edina, MN: Two watersheds are present in the city of Edina; Minnehaha Creek drains the NE section of the city; Nine Mile Creek drains the SW section of the city. Lake Cornelia, which is the object of this paper, is located within Nine Mile Creek Watershed. The lake is 50 acres in surface area and drains 975 acres, including the lake. 66th street divides the lake into a 19-acre north basin and 31-acre south basin. The average depth of the north basin is 1 m and the south basin 1.3 m. (Lake Cornelia UAA, 2010) Figure 2 – Lake Cornelia Coring Location: Lake Cornelia bathymetric map showing the coring location in the north basin of Lake Cornelia. The estimated depth at the coring site was 1.2 m. A Zodiac raft and floating pontoon were used to gain access to the coring site. Lake Cornelia Page 8 Figure 3 – Lake Cornelia deep core: Image of the Livingston core from the north basin of Lake Cornelia, Edina, MN with physical description. Orange line marks the transition from fine lake sediment to plant-rich soil. Note: The first 20 cm was water or water/sediment interface that was lost on extrusion. Impenetrable hard pack sediment was reached at 80 cm. 20-35cm Very fine grain, silty, no plants, gray brown, high water content 35-40cm Very fine grain, silty, grey brown in color, high water content, few very small roots visible 40-44cm Higher root content 44-46.5cm Drier, higher root density, transition boundary between fine grain silty and courser spongy 46.5-58cm Spongy texture (continues to end), warmer brown, high root density 58-60.5cm Contains fine lighter colored grains (quartz?), soil? Lighter brown, high root density 60.5-67.5cm Darker brown, progressively high root density 67.5-70.5cm Contains fine lighter colored grains, soil? Lighter brown, high root density 70.5-72cm Warm brown, high root density 72-80cm Blackish brown, high root density Lake Cornelia Page 9 Figure 4 - Carbon and Nitrogen Analysis: Percent Nitrogen, Percent Total Organic Carbon and CN Ratio for sediment samples from both the Glew sediment core (shown in blue) and the Livingston deep core (shown in orange) from the north basin of Lake Cornelia. The two cores were taken on different days. The two coring sites were near each other but not identical in location. The overlap was estimated based on correlating values in CN data as well as XRF data. Lake Cornelia Page 10 0 5 10 15 20 25 30 35 40 45 50 55 60 65 70 75 80 0 50 100 150 Distance below water/sediment interface (cm)Copper (ppm) Lake Cornelia 0 5 10 15 20 25 30 35 40 45 50 55 60 65 70 75 80 50 250 450 Zinc (ppm) Lake Cornelia 0 5 10 15 20 25 30 35 40 45 50 55 60 65 70 75 80 0 100 200 Lead (ppm) Lake Cornelia 0 5 10 15 20 25 30 35 40 45 50 55 60 65 70 75 80 0 25 50 Arsenic (ppm) Lake Cornelia Figure 5a – X-ray Diffraction Analysis: Concentration of copper, zinc, lead and arsenic (in ppm) from samples from the Glew sediment core (blue) and the Livingston deep core (orange) from the north basin of Lake Cornelia. Table 1 – Sediment Quality Targets (SQT): The Minnesota Pollution Control Agency (MPCA) sediment quality targets (SQTs) determine the sediment quality of a lake. Level I SQT is the contaminant concentrations below which harmful effects on benthic invertebrates are unlikely to be observed. Level II SQT is the contaminant concentrations above which harmful effects on benthic invertebrates are likely to be observed. The level 1 and level 2 SQTs for copper, zinc, lead and arsenic are show in the table at left. These limits are also represented in Figure 5a above. The green shaded area represents concentrations below the level 1 limit and the pink shaded area represents concentrations above the level 2 limit. The four listed elements each currently exceed the level 1 target. The concentrations of both copper and zinc in the uppermost surface sediment are approaching the level 2 SQT and their concentrations are increasing with time. Both lead and arsenic reached their peak at 40± cm downcore and are now below the level 2 target. (Crane and Hennes, 2016) Lake Cornelia Page 11 Figure 5b – X-ray Diffraction Analysis: Concentration of calcium, titanium, chromium and iron are given in ppm for sediment samples from both the Glew sediment core (blue) and the Livingston deep core (orange) from the north basin of Lake Cornelia. Synchronous fluctuations deep in the core could indicate response to changing redox conditions in a wetland environment. Lake Cornelia Page 12 Figure 6. Diatom analysis using microscope imagery of smear slides. Images A-C show planktonic diatoms found in the fine sediment segment of the Livingston core. Images D-G show long needle-like sponges as well as plant material found in drier, plant-rich segment of core. Blown up diatom from Image C strongly resembles Navicula trivalis (Image H), a nutrient loving diatom found in Maine lakes. (Danielson, 2009) H Lake Cornelia Page 13 6 6 Palmer (1965) defines drought impacts as follows: Mild drought: Some of the native vegetation almost ceases to grow. Moderate drought: The least drought-resistant members of the native plant community begin to die and the more xerophytic varieties start to take their place Severe drought: Only the most xerophytic varieties of native vegetation continue to grow. And vegetal cover decreases. Extreme drought: Drought-resistant varieties gradually give way to open cover. More and more bare soil is exposed. The red line (LOESS: LOcally wEighted Scatter-plot Smoother) is a method for fitting a smooth curve between two variables or fitting a smooth surface between an outcome Figure 7: History of droughts in Minnesota from 1900-2018 as measured by the Palmer Drought Severity Index (PDSI). PDSI uses temperature and precipitation data to estimate relative dryness. The scale ranges from -10 (dry) to +10 (wet) where -2 is moderate drought, -3 is severe drought, and -4 is extreme drought.6 Minnesota experienced a moderate to extreme drought from 1929-1941, which peaked in the summer of 1934 with a PSDI score of -7.79. This was preceded by a shorter duration drought from 1920-1924. Historical aerial photographs taken over Edina in 1930 and 1940 (Figure 8) show that Lake Cornelia was dry during this period. Image taken from https://www.ncdc.noaa.gov/cag/statewide/time-series/21/pdsi/all/1/1895-2018 accessed 8 Dec 2018 h Lake Cornelia Page 14 Figure 8 - The Dry Years: Historical photos of Lake Cornelia taken in 1930 (top) and 1940 (bottom) show an absence of water in Lake Cornelia during the drought of 1929-1941. 66th street bisects the photos from east to west. France Avenue appears as a country road at the right of the photos. The area in the center of the photographs, which was Lake Cornelia, appears to be dry. Swim Pool and Point of France Ponds which appear in the upper right quadrant of the 1940’s photo appear water filled. (Edina Water Resources webpage) Lake Cornelia Page 15 Figure 9 – The Transition to Urban: Historical photos of Lake Cornelia taken in 1950 (top) and 1960 (bottom) during the period of rapid urbanization of the area. The northwesternmost section of Southdale Shopping Center’s parking lot appears in the lower right of both photos. Southdale was opened in 1956. A housing development which was started at the same time as the mall, is largely completed by the 1960 photo. County Road 62 is captured during its construction in the bottom photo. Lake Cornelia, is once again filled with water. (Edina Water Resources webpage) Lake Cornelia Page 16 Figure 10 – Current: Photo of Lake Cornelia taken in 2012 (top) showing the current parks, residential/commercial areas and streets/highways. The bottom photo shows the network of storm sewers that feed into the lake. Lake Cornelia’s north and south basin are connected beneath 66th street by a 15” pipe along the west edge, a 12” culvert in the center and a 12” pipe along the eastern edge. (Edina Water Resources webpage) Water generally flows from the north basin to the south basin. The water level in both basins is controlled by a weir which is located at the southernmost tip of the south basin. (Edina Water Resources webpage and Lake Cornelia UAA, 2010) Lake Cornelia Page 17 Figure 11 – Lake Cornelia Watershed 1901 vs Current: The background is a plat map from 1901, with the original Lake Cornelia watershed outlined in red and the current watershed outlined in purple. Lake Cornelia is in the center of the lower section of the map. Nine Mile Creek running NW to SE across the lower left quadrant of the map and Minnehaha Creed running across the upper right quadrant of the map. The natural state of the land between these two creeks and surrounding Lake Cornelia was wetland. Over time, people drained the wetland and manipulated the watershed by adding storm sewers to control the urban runoff into the lake, permanently shifting the lake’s watershed. (Figure courtesy of Gael Geber from NMCWD) Lake Cornelia Page 18 REFERENCES 1. Copper Development Association, Inc. webpage; Copper in Brake Pads. https://www.copper.org/environment/impact/copper-brake.html accessed 4 Dec 2018. 2. Crane, J.L. and S. Hennes. 2016. Ambient sediment quality conditions in Minnesota. Environmental Analysis and Outcomes Division, Minnesota Pollution Control Agency, St. Paul, MN. MPCA Document Number tdr-g. 19pg. https://www.pca.state.mn.us/water/sediment-studies-twin-cities-and- statewidesediment-investigations 3. Danielson, T., J. 2009. Protocols for Calculating the Diatom Total Phosphorus index (DTPI) and Diatom Total Nitrogen Index (DTNI) for Wadeable Streams and Rivers. Department of Environmental Protection, State of Maine Bureau of Land and Water Quality. Doc. Num.: DEPLW-0970A. https://www.maine.gov/dep/water/nutrient-criteria/sop_dtpi_dtni.pdf 4. Donigan, Jr. A.S. and Bicknell, B.R. 2007. Modeling the Contribution of Copper from Brake Pad Wear Debris to the San Francisco Bay. Prepared by Aqua Terra Consultants for Association of Bay Area Governments and California Department of Transportation. 62pg. https://www.researchgate.net/publication/237708843_Modeling_the_Contribution_of_Copper_fro m_Brake_Pad_Wear_Debris_to_the_San_Francisco_Bay 5. Edina Water Resources; Edina Engineering Resources webpage. http://edinagis.maps.arcgis.com/apps/webappviewer/index.html?id=aeb57968722e476f9b6ef2b 86d9326b8 accessed 6 Nov 2018. 6. Fowler, T. 2008. A Brief History of Lead Regulations; Science Progress. 21 Oct 2008. https://scienceprogress.org/2008/10/a-brief-history-of-lead-regulation/ 7. Golding, S. 2008. Suggested Practices to Reduce Zinc Concentrations in Industrial Stormwater Discharges. Environmental Assessment Program for Water Quality Program. State of Washington, Department of Ecology Publication Number 08-10-025. https://fortress.wa.gov/ecy/publications/documents/0810025.pdf 8. Hesterman, Paul D., 1988. From Settlement to Suburb: The History of Edina, MN. Burgess Publishing. 104 pgs. 9. Kieffer, J. M. 2018. City of Edina 2018 Comprehensive Water Resources Management Plan, July 2018. Prepared by Barr Engineering. 425pgs. https://www.edinamn.gov/DocumentCenter/View/5146/Edina_2018_CWRMP_July- 2018_Final?bidId= 10. Kojetin, B. Personal conversation. 25 Nov 2018 11. Lake Cornelia Use Attainability Analysis, Updated Draft. January 2010. Prepared for Nine Mile Creek Watershed District by Barr Engineering. 246pgs. https://www.ninemilecreek.org/wp- content/uploads/UAA_LakeCornelia_January2010.pdf 12. MN Department of Health webpage. Consumption Advice for Bodies of Water Where Fish have been Tested: Fish Consumption Guidelines for Women Who Are or May Become Pregnant, and Children under Age 15, Lakes. 184pgs. http://www.health.state.mn.us/divs/eh/fish/eating/specpoplakes.pdf accessed 13 Dec 2018 13. Minnesota Pollution Control Agency webpage: 2018 Impaired Waters List. http://www.pca.state.mn.us/water/minnesotas-impaired-waters-list accessed 14 Nov 2018 14. Morse-Kahn, D. The Nine Mile Creek Watershed District: Preserving Heritage & Environment. Nine Mile Creek Watershed District, 2009. Lake Cornelia Page 19 15. Nine Mile Creek Biological Stressor Identification. November 2010. Prepared for Minnesota Pollution Control Agency. Barr Engineering Report. 102 pages. https://www.pca.state.mn.us/sites/default/files/wq-ws5-07020012a.pdf 16. NOAA National Centers for Environmental Information webpage. Climate Monitoring>Climate at a Glance. https://www.ncdc.noaa.gov/cag/statewide/time-series/21/pdsi/all/1/1895-2018 Accessed 8 Dec 2018 17. Palmer, W., C. 1965. Meteorological Drought. U.S. Weather Bureau, Office of Climatology Research Paper No. 45, 65pgs. https://www.ncdc.noaa.gov/temp-and- precip/drought/docs/palmer.pdf 18. St. Louis Park Historical Society webpage: Southdale Shopping Center. slphistory.org/Southdale/ accessed 14 Nov 2018 19. Steil, M., 2016. Summer of 1936 brought heat and grief. MPR News. July 11, 2016. https://www.mprnews.org/story/2016/07/10/summer-1936-brought-heat-and-grief accessed 24 Nov 2018 20. Walsh, P., 2017. Dangerous toxic algae levels put Edina lake off limits to people, pets. Star Tribune. August 21, 2017. From <http://www.startribune.com/dangerous-toxic-algae-levels-put-edina-lake- off-limits-to-people-pets/441287203/> Accessed 15 Oct 2018.