| Summary: | This report presents a trend analysis conducted for nutrient-related water quality parameters collected at monitoring stations located on the Fox River main stem and tributariesand compiled and maintained in a database, FoxDB. An exploratory data analysis (EDA) was performed on a total of 141 water quality parameters across the 18 monitoring stations to summarize and extract the characteristics of the water quality data. Based on the EDA analysis, the Seasonal Kendall Test (SKT) for trends was selected as the core analysis method, and the EnvStats software R-package was used to perform the water quality trend analysis and EDA. A suite of procedures and workflows that use the EnvStats library of codes for the analysis were written using R programing language to extract selected water quality data from the FoxDB (i.e., the water quality database for the Fox River watershed) and perform the analysis. In addition to the nonparametric analysis using the SKT method, trend analyses of water quality concentration and fluxes (loads) were conducted for one Fox River main stem and two tributary monitoring stations that have not only long-term concentration data, but also the corresponding continuous daily discharge data. A total of 19 parametric models using concentration and flow data across the three stations were developed using the Weighted Regression on Discharge, Time, and Season (WRTDS) method for estimating trends in flow-normalized concentration and fluxes. For all monitoring stations, the SKT trend analysis generally showed that most of the nutrient-related water quality parameters exhibited either a decreasing or no trend across all seasons. No upward annual trend was exhibited for organic nitrogen (Org-N), ammonia nitrogen, total suspended solids (TSS), or chlorophyll-A (CHL-A) at any of the monitoring stations. At the most downstream station on the main stem, the Fox River at Yorkville, no increasing trend wasdetected, with most of the water quality parameters showing a decreasing trend across all seasons. Most of the upward trend was detected for dissolved phosphorus (DP), particularly in spring and summer months. In comparison, total phosphorus (TP) showed an increasing annual trend only for Poplar Creek near the Mouth-Elgin station. For more than half of the stations, the pH showed a downward or no trend. In the case of pH, an upward or downward trend from the median, which is within the pH limits for freshwater, would indicate a declining water quality. All remaining water quality parameters exhibited decreasing longitudinal trends downstream of the Fox River at Algonquin, indicating improvement of the river’s water quality, except for dissolved oxygen (DO), which rather implies a declining water quality trend. The results of the trend analysis conducted using the WRTDS method generally indicate that flow-normalized concentration and fluxes (loads) of most water quality parameters decreased across all seasons from 2006 to 2016 for the Fox River at Montgomery, which is the only station in the main stem with the required concentration and flow data. A few exceptions were concentration and fluxes of total suspended solids (TSS) in spring and chlorophyll A (CHL-A) in summer, which showed increasing trends. If not in the percentage amount, the flux and concentration trends are largely similar for this station (i.e., they are in the same downward or upward direction). The only difference observed was between the spring ammonia nitrogen (NH3-N) concentration and its corresponding flux, which showed opposing trends, indicating that concentration trends are not necessarily informative of flux trends. Large decreases in summer DP, NH3-N, and nitrate nitrogen (NO3-N); winter TP, TSS, and CHL-A; and spring for DO, Org-N, and total kjeldahl nitrogen (TKN) concentrations were obtained for the Fox River at Montgomery station. A decreasing trend in concentration across all seasons, unlike for DO, is VII indicative of an improving water quality trend. In comparison with other water quality parameters, flow-normalized fluxes of TP and DP also appeared to have larger decreases across all seasons between 2006 and 2016. Similar downward trends of nitrate nitrogen (NO3-N) fluxes were obtained in summer and fall seasons. For the two tributaries, Blackberry Creek at Rt. 47 and Poplar Creek near Mouth-Elgin, most of the water quality concentration and fluxes showed larger upward trends with a few exceptions. The NH3-N concentration exhibited the largest annual and seasonal increasing trends at both stations. Concentrations of TP, DP, and DO showed decreasing annual and seasonal trends for Blackberry Creek at Rt. 47, except in fall for DO and in summer for TP and DO concentrations. For Poplar Creek near Mouth-Elgin, the DP and DO concentrations showed improving water quality trends across all seasons. The flow-normalized DP and TKN fluxes exhibited decreasing annual and seasonal trends for Poplar Creek near Mouth-Elgin and Blackberry Creek at Rt. 47, respectively. T he seasonal concentration trends largely conform to the annual trends for all three monitoring stations. In addition to water quality trends, flow durations and trends of selected streamflow statistics including mean, 7-day minimum, and 1-day maximum flows are calculated to evaluate their changes through the years as they relate to water quality. The flow durations allow characterizing the ranges of flows in the river that are common or extreme during an entire year or season. The results indicate that the highest and lowest flow variability occurred in summer and spring seasons, respectively. The mean flow provides information about the central tendency of the multiyear hydrologic variability, whereas the minimum and maximum flow trends may explain part of the increase or decrease in constituent concentrations and fluxes. However, to explicitly attribute the change in water quality trends to some changes in hydrologic factors, the extent of other potential factors influencing water quality, such as conservation efforts, land use changes, etc., also need to be examined. Between 2006 and 2016, the mean and 7-day minimum flows exhibited an increasing trend with varying magnitudes across all seasons except for the spring 7-day minimum flow. Generally, the annual and seasonal 7-day minimum flows seem to show large increases during the period of analysis. The annual and seasonal 1-day maximum flows show increasing trends for Blackberry Creek at Rt. 47. For Poplar Creek near Mouth-Elgin, however, the 1-day maximum flow exhibited a decreasing trend in winter, spring, and fall seasons, whereas its annual and summer values had increased. Fox River Study Group published or submitted for publication is peer reviewed Open
|
|---|