Data Analysis

Sample Data Analysis

Sampling Requirements

As stated above, existing TMDLs may only require that the median monthly concentration, or

perhaps a monthly average, be reported. This value is then compared with the target to assess

compliance. These calculations are straightforward. Simple spreadsheets can be used to repre-sent

the data.

It is also important to have solid "metadata" recorded and clearly accessible for all sites and

parameters: the weather and other environmental conditions under which the samples were

collected; sample handling, storage, and preparation; the types of analyses conducted; specific

methods of analysis; and the quality assurance/quality control measures that were taken.

EcoWatch - Data Analysis Software - Visualization Tool

Continuous monitoring, and even discrete sampling, can generate huge volumes of data. Without

a software package that allows the easy manipulation of data – the ability to "zoom in" or "zoom

out" on data along a timeline, or to explore correlations between parameters – it’s difficult to

analyze your sample results. In short, good manipulation and analysis tools help narrow down the

sources of pollution and focus on solutions.

Many users are also finding that data analysis software makes a great visualization tool to help

stakeholders understand how their portion of the landscape relates to water quality. It is impor-tant

that landowners have a means of visualizing their land area or outfall in relation to the rest of

the water body and watershed in which they are situated. Graphing features in data analysis

software provide excellent tools to help landowners and the general public visualize water quality

issues facing their streams and lakes, as Wyoming conservation districts have been doing with

great success.

Following are some examples of screens from YSI EcoWatch for Windows PC software. The

data provide background information on the water quality of a 20-mile stream in southwestern

Ohio that forms the "spine" of a wetlands that is being restored and preserved.

For an initial study, a YSI 6000UPG multi-parameter water quality monitoring instrument called

a "sonde" was set up to log the following parameters to internal memory every 30 minutes:

temperature, dissolved oxygen, conductivity, pH, ORP (Oxygen Reduction Potential, or Redox),

and turbidity. The sonde was then deployed in the stream at a depth of 2 feet and left for 60 days

before recovery and upload of the information to EcoWatch. The data as plotted in EcoWatch

are shown in Figure 3.

For most of the study, a diurnal pattern in dissolved oxygen, pH, and ORP is observed: a common

cycle which is due to the activity of plants and phytoplankton which produce oxygen during the

daylight hours through photosynthesis and consume it at night through respiration. Note that

several rainfall events occurred during the deployment, evidenced by concurrent decreases in

conductivity and increases in turbidity (a standard pattern for the effect of rainfall on surface

water). The final event corresponded to approximately 3 inches of rainfall during a 24-hour

period and not only resulted in extreme perturbation of the turbidity/conductivity reading, but

also in the "washing" out or dilution of the biological diurnal cycle in oxygen, pH, and ORP.

Figure 3

It is unlikely that the two key aspects of the site – diurnal cycles in key parameters and turbidity/

conductivity perturbations with rainfall – would have been identified and quantified with spot

sampling studies alone.

Fine-tuning the data display to focus on key parameters is an important way to avoid information

overload. The above screen shows all the data collected from 10/30/98 through 12/30/98, dis-playing

turbidity, pH, DO, specific conductance, ORP and temperature. You can add additional

parameters being monitored, or remove some parameters to focus on a specific question. For

instance, figure 4 graphs only turbidity and conductivity.

Figure 4

Manual scaling provides a clearer view of the diurnal turbidity cycles on figure 5 by allowing us

to "zoom in" to a smaller range of values on the y-axis.

Figure 5

Figure 6

A few clicks of the mouse allow you to view the data at any time interval. Instead of looking at

all 60 days of data as shown in figures 3 through 5, you could zero in for a detailed look at just

one week, as in figure 6.

The results of short-time-span study can be

fascinating, even mysterious. Rather than

showing the turbidity data over the 60-day time

span, figure 6 charts turbidity over a one-week

timeframe – easily accomplished with a few

clicks of the mouse. Note the nighttime spikes

in turbidity. Why is there a diurnal turbidity

cycle? Local experts suspect that nocturnal

animal activity is causing a rise in the turbidity

levels.

Is this river resident’s nocturnal activity

responsible for diurnal turbidity peaks?

Figure 7

Figure 7 shows DO and pH in separate graphs, each plotted against time on the X-axis. Compar-ing

the graphs, it is easy to see that the two parameters rise and fall in step with each other.

Figure 8

The parameters on the X- and Y-axes can easily be changed too, providing alternate ways to view

data. In figure 8, EcoWatch allowed us to view the same data with pH on the x-axis and DO on

the y-axis. The resulting graph shows a strong, positive correlation between the two factors.

Further analysis could quantify the slope of the line to measure how closely correlated the two

parameters are. This function could be especially helpful in exploring correlations between other

parameters, answering questions like, "What is the relationship between turbidity and TSS?"

With EcoWatch and off-the-shelf spreadsheet software, you can import, combine, and manipulate

data from a variety of sources. For example, data from a flow meter can be imported into the

spreadsheet along with the data from a multi-parameter instrument built by any manufacturer.

Loading the spreadsheet into EcoWatch allows you to manipulate the data to explore specific

relationships. Data from EcoWatch can also be exported into a database for further combination

and manipulation with data from other sources.

Additional EcoWatch features include:

w Easy exporting of data to a spreadsheet or to BASINS

w Instantaneous statistics such as minimum, maximum, mean and standard deviation

w Easy merging of data to accommodate long-term studies

w Simple connection to a wide variety of modems, which allows viewing of real-time data

from your office

Figure 9

Finally, EcoWatch also offers a user-configured conversion of turbidity into total suspended

solids. Based on data points established with your calibration and laboratory work, the software

will automatically make the calculations. Figure 9 shows the screen for the turbidity and TSS

correlation table where the TSS data points are added.

Figure 10

Figure 10 displays the graphs of turbidity and TSS derived from the data entered into the correla-tion

table.

(Note: The exact type of correlation between turbidity and TSS may be quite site-specific; this

information is not meant to imply that there is any universal correlation between turbidity and

TSS.

Imagine trying to process and comprehend data for several parameters, from sampling sites

across a large watershed. Without a good software package, it would be time-consuming and

difficult to effectively analyze and evaluate water quality as well as make decisions about solu-tions

to your water quality problems. Fortunately, good software exists – and even offers the

flexibility of exporting data into sophisticated modeling programs such as BASINS, which is

discussed in the next section.