Tale of Two Storms: Impact of Extreme Rain Events on the Biogeochemistry of Lake Superior

Abstract

Climate change is expected to profoundly affect the Great Lakes region of North America. An increase in intensity and frequency of rain events is anticipated to deliver more runoff and to increase riverine inputs to Lake Superior's ecosystem. The effects of these changes on key biogeochemical parameters were analyzed by coupling satellite data, water column sensor profiles, and discrete surface-water sampling after two "500-year" flood events in the Lake Superior basin. This study provides both a spatial and a temporal sense of how plumes interacted within the ecosystem. We also determined the significant differences in water quality parameters for plume versus non-plume waters. These two plumes were important for delivery of nutrients, with variable transport of sediments and colored dissolved organic matter (CDOM). Data from the 2012 storm event showed a significant input of total nitrogen (TN), total phosphorous (TP) and CDOM to the system. In the 2016 storm event, carbon cycling parameters (acidity, total inorganic carbon (TIC), and dissolved organic carbon (DOC), and ammonia levels were elevated within the plume. In neither storm event was there a significant difference in chlorophyll a between plume and non-plume waters during our sampling cruises. These two plume events were similar in amount of precipitation, but their effect on the biogeochemistry of Lake Superior varied due to differences in the watersheds where the rain fell. The studied plume events were dynamic, changing with currents, winds and the settling of suspended sediments.

Figure 1.

Total rainfall amounts for each storm event in inches with the watershed regions for riverine input to far-western Lake Superior outlined in grey. Dots represent rain gauge data collection locations. a. 2012 storm with St. Louis watershed labeled. b. 2016 storm with Nemadji upper and lower watershed and Chequamegon Bay labeled.

Figure 2.

a. Map of Lake Superior (from Google Earth) showing the location of the NOAA lake level gauges whose data is shown in b & c. Inset to a shows all the Laurentian Great Lakes. Fig. 2.b & c. Water level rise due to storm events based upon the average lake level height data of NOAA gauges 9099064 (Duluth, MN) and 9099004 (Point Iroquois, MI) collected every hour; b. shows the 2012 storm event; C. shows the 2016 event. Dashed horizontal lines show the low and high for the respective year averaged for the two gauges. The vertical line shows the onset of the extreme rain event studied.

Figure 3

Stars show locations of plume and triangles show non-plume sites of ship-based sampling sites overlaid on K_d (490) values in SeaDAS software. Red indicates K_d (490) values equal or over 0.41 m⁻¹ (plume), Blue indicates K_d (490) values under 0.41 m⁻¹ (non-plume). White show areas were no data was available either due to extensive loading which caused a breakdown in satellite optical measurements, or due to cloud cover. Grey indicates land as determined from a preloaded mask in SeaDAS. Pixel size 0 .75 km x 0.75 km at nadir

Figure 4

Stars show locations of plume, triangles show non-plume, and squares indicate unclassified sites of ship-based sampling sites overlaid on K_d (490) values in SeaDAS software. Red indicates K_d (490) values equal or over 0.41 m⁻¹ (plume), Blue indicates K_d (490) values under 0.41 m⁻¹ (non-plume). White show areas were no data was available mainly due to extensive loading, or due to cloud cover. Grey indicates land as determined from preloaded mask in SeaDAS. Pixel size 0.75 × 0.75 at nadir.

Figure 5

Image shows the Dominate Soil Order in the Sub-watersheds of Western Lake Superior, Areas displayed inside the watershed in grey are areas that have not been surveyed yet. The image was produced using ArcGIS Pro.

Figure 6

Image shows the varied land use in each of the sub-watersheds of Western Arm of Lake Superior. 6a. focuses on the St. Louis watershed, and 6b. focuses on the watersheds that are on the Wisconsin shoreline. Image was produced in Arc GIS Pro.