As mentioned previously for each storm event, three samples were collected, sample bag 1, sample bag 2 and the dry bag. The purpose of having two sample bags was to provide an average phosphorus reading for a sample area that would take into consideration variations within the sample protocol and location. The possible influences included bugs, tree pollen and other natural effects. The following comparison graphs show the concentrations Bag 1 and Bag 2 per sampling location.
Graph 1
Graph 2
Graph 3
Graph 4
The purpose of having the dry bag was to collect local atmospheric deposition and to use that phosphorus value as a control blank. Prior to the placement of any collection bags a field control analysis was conducted on the bags to rule out the addition of any phosphorus from initial bag contamination. Based on the average analysis of the dry bags each sample location was ranked for local atmospheric deposition. The following table provides that ranking per location.
Location |
Average Dry Bag |
Rank |
30 lakes |
0.275 |
1 |
Shell Rock Bartusek House |
0.081 |
2 |
Red Lake |
0.055 |
3 |
Shell Rock Court House |
0.036 |
4 |
Table 6
Analysis of the data, including the HYSPLIT graphs and GIS tracking, further solidified the hypothesis that the increase seen in phosphorus levels were coming from atmospheric loading. To determine exactly where geographically the loading was coming from, the GIS tracking data was modified to show when the storm event air mass was at or below 500 meters during the 120 hour backwards trajectory. Statistical analysis of the storm event elevations versus the total phosphorus concentrations revealed a high correlation. Basically indicating that the lower the elevation of the storm event, the higher the total phosphorus concentration was for the overall storm event. Correlation coefficients for both the extreme and low loadings were determined for all three Watershed Districts. Some data points were omitted because of they did not represent the commonality of the rest of the data. There are two correlation graphs shown per location. The first correlation graph shows all extreme and low data points. The second is a correlation graph with the omitted data points. The following image shows the storm event track of the omitted data along with image maps showing areas where the air mass was at or below the 500 meters (indicated in red) and the areas that the air mass remained greater than 500 meters (indicated in shades of purple) for both the extreme and high precipitation events.
30 Lakes
Correlation Graph 5
Omitted Correlation Graph 6
Omitted tracking Image 114
Extreme events Image 115
High events Image 116
Red Lakes
Correlation Graph 7
Omitted Correlation NA
Omitted tracking NA
Extreme events Image 117
High events Image 118
Shell Rock Court House
Correlation Graph 8
Omitted Correlation NA
Omitted tracking NA
Extreme events NA
High events Image 119
Shell Rock Bartusek House
Correlation Graph 9
Omitted Correlation Graph 10
Omitted tracking Image 120
Extreme events Image 121
High events Image 122
30 Lakes Correlation
    
Correlation Coefficient |
-0.67625 |
Red circles indicate the omitted values on the next graph
Graph 5
30 Lakes Correlation – Omitted Values

Correlation Coefficient |
-0.87968 |
Graph 6
30 Lakes Correlation – Omitted Values Tracking

Image 114
30 Lakes Extreme

Image 115
30 Lakes High

Image 116
Red Lake Correlation

Correlation Coefficient |
-1 |
Graph 7
Red Lake Extreme

Image 117
Red Lake High

Image 118
Shell Rock Court House Correlation

Not enough data to run a correlation
Graph 8
Shell Rock Court House High

Image 119
Shell Rock Bartusek House Correlation
 
Correlation Coefficient |
-0.67335 |
Red circles indicate the omitted values on the next graph
Graph 9
Shell Rock Bartusek House Correlation – Omitted Values

Correlation Coefficient |
-0.71995 |
Graph 10
Shell Rock Bartusek House Correlation – Omitted Values Tracking

Image 120
Shell Rock Bartusek House Extreme

Image 121
Shell Rock Bartusek House High

Image 122
To make it easier to identify, the air masses less than 500 m, were mapped and all other elevations were eliminated. This data was then mapped out by sampling location as shown by the in the images below.
30 Lakes

Image 123
Red Lake

Image 124
Shell Rock Court House

Image 125
Shell Rock Bartusek House

Image 126
|