Laserfiche WebLink
Because the data showed no obvious trends decipherable from the raw <br /> data, and because the AHRs indicated poor quality water with <br /> respect to TDS but no alarming changes, I elected to take a more <br /> rigorous statistical approach for the evaluation. By evaluating <br /> the following parameters at Station #20, I drew the conclusions <br /> which I will outline: <br /> Date of sampling <br /> Rate of Flow <br /> Conductivity <br /> Total Suspended Solids <br /> Total Dissolved Solids <br /> PROCEDURE <br /> 1. Plot Conductivity, TSS, and TDS relative to time. <br /> Conductivity and TDS in this system and in many others behave <br /> in very similar ways. Both are an indirect measure of the <br /> dissolved chemical components in the water than may affect <br /> aquatic life. TSS is a direct measure of the physical <br /> (sediment) component that may affect aquatic life, and is an <br /> indirect, partial measure of the dissolved component. <br /> 2 . Evaluate the relationships between flow and Cond, TSS, and TDS <br /> by calculating the fit of the data about a regression line and <br /> calculating the standard deviation about that line. <br /> Curve-fitting is a way of identifying anomalous values that <br /> may not appear anomalous when compared with, say, the mean of <br /> the entire data set. For instance, for any given rate of <br /> flow, there is a fairly narrow band of variation in TDS, and <br /> the TDS:Flow relationship is an inverse relationship; low flow <br /> correlates with high TDS and high flow correlates best with <br /> low TDS. A higher than average TDS value in the mid- to high- <br /> flow range may fall within one standard deviation of the mean <br /> of the entire TDS range, regardless of flow, but may still <br /> indicate some perturbation in the system. <br /> 3 . Examine Cond, TSS, and TDS values that are more than one <br /> standard deviation above the mean of the regression curve <br /> (line) and determine whether these outliers bear any relation <br /> to time of sampling. <br /> The question is whether recently-collected samples record <br /> perturbations in the system, and whether such perturbations <br /> can be linked to on-site activities. <br /> RESULTS <br /> The results are discussed in terms of flow, conductivity (and TDS) , <br /> TSS, and date of sampling. All of the following discussion refers <br />