50% Off Our Lifetime Plan! Ends Soon.Learn more

Developing a Water Quality Model

3.7 based on 7 ratings

Updated on Jan 29, 2014

Problem

My problem is to correlate Gravel Run and Centreville Landing in terms of pH, dissolved oxygen, turbidity/clarity and air and water temperature. My project will help to determine the relationship between Gravel Run and Centerville Landing, or to see if the data I have been collecting moves together. It will be the first step in building a computer model of an estuary.

Background Information

Variables: Temperature - The temperature of the water is how hot or cold the water is. We measure the air and water temperature at each site using thermometers calibrated against NBS certified thermometers. The temperature of the water is especially important to calculate the dissolved oxygen saturation level.

Dissolved Oxygen - Dissolved Oxygen is essential for most organisms that live in the stream. Several things would have an effect on the DO of Centreville Landing. The first is temperature. When the temperature is warmer, the water can hold less oxygen. When the level of oxygen goes below 3 ppm it is harmful and deadly to most all life. When the temperature is colder the water can hold more oxygen, causing a higher reading. It is better to have it this way. Another factor for dissolved oxygen would be decomposition. This is because as things decompose in the stream, oxygen is used; thus there is less DO.

Something else that affects dissolved oxygen is respiration. Animals use oxygen to live. When testing for dissolved oxygen, it sometimes depends on what time of day you do it at. Levels are always the highest in early morning because plants have stopped creating oxygen at night (the sun goes away), but organisms have been using it all night long. Thus in the morning most of the oxygen produced by plants the day before is used and animals start dying.

Nitrates - Nitrates are the most common form of nitrogen. They are essential for growth; though too much from waste, decomposition, and fertilizer can be a problem. Nitrates are measured in parts per million or ppm. From past monitoring experience, I have known Gravel Run to have higher nitrates. This could be for many reasons, some being excess farm run-off, lawn fertilizers and sewage. Centreville has not had high readings since I started monitoring.

pH - The pH in water is how acidic or basic (alkaline) it is. It is measured on a scale from 0-14. Measurements from 0-7 show how acidic the water is. Zero is the most acidic and seven is neutral, or pure water. A measurement of pH from 7- 14 show how basic the water is, fourteen being the most basic. Some factors that might affect the pH are rainfall (if there is a lot of acid rain) and organism activity.

Turbidity/Clarity - Turbidity is how dirty or cloudy the water is. It is caused by bits of clay, silt, plankton and sewage suspended in water. You can use a turbidity test kit, which measures in JTU’s (Jackson Turbidity Units). Clarity is how clear the water is, measured with a secchi disk. We measure the clarity at Centreville Landing because the water is deep enough to use a secchi disk.

The Sites: Gravel Run is a freshwater stream that flows into Centreville Landing from the center of Queen Anne’s County.

Estuaries: Centreville Landing (on the Corsica River) is an estuary, where fresh water from Gravel Run (upstream) and brackish water from the Chester River (tidal water) mix.

The effects of the salt water on fresh water: When fresh and salt water mix together in an estuary, they sometimes form layers. This is called stratification. Because the density of the salt water is greater than that of the fresh, it will lie on the bottom. Physical forces such as tides, wind, waves and river runoff, contribute to the mixing of estuaries. In areas, like the Chesapeake Bay, with a smaller river flow and larger tides, more mixing takes place and less layering. The different variables from the Bay and Gravel Run determine the water quality at Centreville Landing. Most of the nitrates come from the fresh water. This is because there is more decomposition of plants and algae, more runoff from farmers’ fields and also animal and industrial waste. The turbidity is also greater in freshwater areas because of more algae, run-off, and dirt. The ocean water is more diluted, and most dirt does not get mixed up as much as it does in freshwater. All of the salinity is found in the ocean, caused by dissolved salts. The salinity of ocean water is about 35 ppt (parts per thousand). Although freshwater does contain salt, it is usually less than 1 ppt.

Correlation: Correlation is used to determine the relation between two properties. Correlation is between 1 and –1, with zero as no correlation, one as positive and – 1 as negative correlation. When the correlation is found between two sets if numbers, it is seeing how well they move together. When one number of the set goes up does the other go up (positive), does the other number go down (negative) or they don’t move together at all (zero).

Hypothesis

I think that there will be a correlation in the data I collect for Gravel Run and Centreville Landing. The air and water temperatures should move together, when one goes up so does the other (positive correlation). This is probably because the sites are relatively close together and have the same kind of climate. The pH might have close to a correlation of one, because in looking at past data, I can conclude it is almost always around seven. The amount, not changing for either site. The dissolved oxygen levels should also move together because the DO depends on the temperature of the water. If the temperature goes up, the DO should go down, on both sites. The nitrates should also have positive correlation. Since we measure turbidity at Gravel Run, which is how dirty the water is and clarity at Centreville Landing, how clean the water is, there should be negative correlation, when one goes up the other should go down.

Materials

1 of each test kit - Dissolved Oxygen

  • Nitrates
  • pH
  • Turbidity Thermometer

Secchi Disk Computer spreadsheet program

Procedure

  • Monitor the water quality of two sites along the Chester River for a year.
  • Enter the data (temperature, nitrates, dissolved oxygen, pH, and clarity/turbidity) in a spreadsheet.
  • Using the spreadsheet formula to correlate each set of data for each category.
  • Determine the effect of Gravel Run on Centreville Landing

Nitrates Test

  1. Fill a test tube (0124) to the 5-ml line with sample water.
  2. Add one Nitrate #1 Tablet (2799).
  3. Cap and mix until tablet disintegrates.
  4. Add one Nitrate #2 CTA Tablet (NN-3703).
  5. Cap and mix until tablet disintegrates.
  6. Wait 5 minutes.
  7. Insert test tube into Octa-Slide Viewer.
  8. Match sample color to a color standard. Record as ppm Nitrate-Nitrogen.
  9. To convert reading to ppm Nitrate, multiply results by 4.4.

Secchi Disk

  1. Lower Secchi Disk into water until it just disappears.
  2. Read depth in meters from calibrated line.
  3. Raise Secchi Disk until it just appears. Read depth from calibrated line.
  4. Add readings from Steps 1 and 2. Divide by 2.
  5. Record as Secchi Disk Transparency.

Turbidity

  1. Fill one Turbidity Column (0835) to the 50-ml line with the sample water.
  2. If the black dot on the bottom of the tube is not visible when looking down through the column of liquid, pour out a sufficient amount of the test sample so that the tube is filled to the 25-ml line.
  3. Fill the second Turbidity Column (0835) with an amount of turbidity-free water that is equal to the amount of sample being measured. Distilled water is preferred; however, clear tap water may be used. This is the "clear water" tube.
  4. Place the two tubes side by side and note the difference in clarity.
  5. If the black dot is equally clear in both tubes, the turbidity is zero.
  6. If the black dot in the sample tube is less clear, proceed to Step 4.
  7. Shake the Standard Turbidity Reagent (7520) vigorously.
  8. Add 0.5 ml to the "clear water" tube. Use the sitting rod (1114) to stir contents of both tubes to equally distribute turbid particles.
  9. Check for amount of turbidity by looking down through the solution at the black dot. If the turbidity of the sample water is greater than that of the "clear water", continue to add Standard Turbidity Reagent in 0.5 ml increments to the "clear water" tube, mixing after each addition until the turbidity equals that of the sample.
  10. Record total amount of Turbidity Reagent added. Each 0.5-ml addition to the 50-ml size sample is equal to 5 Jackson Turbidity Units (JTU's). If a 25-ml sample size is used, each 0.5-ml addition of the Standard Turbidity Reagent is equal to 10 Jackson Turbidity Units (JTU's).

pH

  1. At each sampling a wide range and narrow range test should be done. If the reading is within the narrow range, this reading will be used. A wide range test will also be done to a) Serve as a check for the narrow range reading and b) To provide a more accurate reading if the narrow range is exceeded (Note: if the narrow range is exceeded, the color will match one of the two extreme readings, and only the wide range reading will indicate whether the reading is out of range or is exactly one of the extreme readings.
  2. Make sure to rinse the sample test tube 3 times with the water to be tested.
  3. Cap the tube to mix water and the indicator; do not use your finger.
  4. Sometimes due to degradation of the indicator, a color will be obtained that is more or less intense than the octet comparator colors. Redoing with slightly more or less drops can change the intensity for a better "match" without changing the hue.

Dissolved Oxygen

  1. Thoroughly rinse the Water Sampling Bottle (0688-DO) with sample water 3 times (or use a sampling device that does this) then fill the bottle. The sample is then collected from one foot below the surface (where permitted by depth).
  2. Tap the sides of the submerged bottle to dislodge any air bubbles clinging to the inside.
  3. Once a satisfactory sample has been collected, proceed immediately with next steps, to "fix" the sample.
  4. Be careful not to introduce air into the sample while adding the reagents. Simply drop the reagents into sample, holding the reagent bottles vertically.
  5. Add 8 drops of Manganous Sulfate Solution (4167) and 8 drops of Alkaline Potassium Iodide Azide (7166).
  6. Cap and mix by inverting several times. A precipitate will form. Allow the precipitate to settle below the shoulder of the bottle before proceeding.
  7. Add 8 drops of Sulfuric Acid, 1:1 (6141WT).
  8. Cap and gently shake until the reagent and the precipitate have dissolved. A clear yellow to brown-orange color will develop, depending on the oxygen content of the sample.
  9. Fill the titration tube (0299) to the 20-ml line with the "fixed" sample and cap.
  10. Fill the Direct Reading Titrator (0377) with Sodium Thiosulfate, 0.025N(4169).
  11. Insert the Titrator into the center hole of the titration tube cap. While gently swirling the tube, slowly press the plunger to titrate until the yellow-brown color is reduced to a very faint yellow. NOTE: If the color of the "fixed" sample is already a very faint yellow, skip to Step 13.
  12. Remove the Titrator and cap. Be careful not to disturb the Titrator plunger, as the titration begun in Step 12 will be continued in Step 14. Add 8 drops of Starch Indicator Solution (4170WT). Sample should turn blue.
  13. Replace the cap and Titrator. Continue titration until the blue color just disappears. Read the test result where the plunger tip meets the scale.
  14. Record as ppm dissolved oxygen.

Data/Graphs

See http://www.qacps.k12.md.us/cms/sci/PROJJEW.HTM#data

Data Discussion

In looking at the correlation’s from the data I found some very surprising things. Only two of the six variables I tested had data that correlate throughout the year. The water temperature had a correlation of .83, which is close to one. The air temperature also was correlated between the two sites. It had a positive correlation of .93. All the other variables showed no correlation at all. T he pH was -.13; the dissolved oxygen had a correlation of -.25, the nitrates -.01, and the turbidity/clarity a correlation of .18. All of these numbers are so close to one (zero correlation) that I considered them basically 0. There is room for human error because a human does each test. I do not consider this much of a factor because anyone who helped me with doing the test would be certified in water quality monitoring.

Analysis and Conclusion by Variable

Air Temperature

The air temperature data for Gravel Run and Centreville Landing had a positive correlation of .83. This means that as one reading was up, so was the other. The air temperature having a positive correlation makes sense because the sites are close together in terms of distance. Since they are near each other, they have the same climate. When it is extremely warm at Gravel Run, it would only make sense for it to be extremely warm at Centreville Landing.

Water Temperature

The water temperature at Gravel Run and Centreville Landing was also correlated. The data had a positive correlation of .93. This correlation also seems to make sense because if the air temperature is below freezing, then the water temperature should be below freezing. The biggest affect of the water temperature would be the air temperature. Since the air temperature was correlated, so should the water.

Dissolved Oxygen

The two sets of data were not at all correlated in terms of dissolved oxygen. There was zero correlation of -.25. I found this very strange because of several reasons. Since the water temperature was correlated, and the temperature directly affects the dissolved oxygen, I thought that they would be correlated. If the water is cooler, than the DO is higher. This is because cooler water holds more oxygen. When I found that the numbers did not correlate, I correlated two other sites on the Chester River in terms of dissolved oxygen. I found that these did have a positive correlation of about .84.

Some reasons that the data could not have correlated would be because of the movement of the water. There is faster moving water at Gravel Run and slower water at Centerville Landing. This would have an effect because faster moving water contains more oxygen than slower moving water. There is more shade at Gravel Run, causing it to be cooler and it does not let as much sun in for photosynthesis to take place. Centerville Landing is in full sun. There is less algae at Gravel Run, and it would produces more oxygen

Nitrates

The nitrate readings from both sites did not correlate either. The two sets of data had a correlation of -.01. The data could not have correlated for several reasons. Gravel Run receives more ground water than Centerville Landing containing more nitrates. The run off, ground water and rain in Centreville Landing is all very diluted by the salt water coming in. The salt water contains very few nitrates, thus diluting it. Another reason that Centreville Landings water and nitrate level are more diluted is because of other small stream that flow into it. There is also more agricultural land around Gravel Run and more run- off from the fields. At Centerville Landing, there are tides and at Gravel Run there aren’t any.

pH

The pH readings for both sites did not correlate either. They had a negative correlation of .13. This also was not what I expected. Some reasons for this could be are there is more ground water coming into Gravel Run, which might mean more acidic readings. At Centreville Landing there is more salinity, causing the water to be more basic. There might be more life at one site, causing more respiration, and when carbon dioxide is combined with water a type of acid is created. This would cause the water to have a lower and more acidic pH.

Turbidity/Clarity

The clarity and turbidity did not correlate either. I thought that there would be a negative correlation because the tests are testing opposite things. There was a positive correlation of .18. Reasons for this difference might be that there is more algae at Centreville Landing and the runoff that increases Gravel Run’s turbidity is more diluted.

Conclusion

When looking at my results for finding the correlation of water quality for two sites on the Chester River, I can come to several conclusions. After I found out the correlation between the two sets of data, I started to find out why the two sets of data did or did not correlate. It is fairly hard to do this because of all the factors that have an effect on the water quality.

In looking at my data when it was all collected, I found many reasons that it could be wrong or not showing an accurate reading. There were some times during the year when it was not possible to monitor the sites, causing some data to be missing. Since sometimes the data was collected for one site but not the other, it is not possible to tell whether there was a correlation for that week, and it also would affect the entire years’ correlation.

When I found out that most of the data that I had thought would correlate did not, I immediately thought my project was bad and that I would never be able to figure out why it happened. When I sat down and actually thought and talked about it with Mr. Radcliffe, I realized I was wrong. I had thought that since the sites were close together and one flowed into the other they would naturally affect each other. I learned during this project that it does not really matter. Each site is so different because of the surroundings (agriculture, marshes, impervious surfaces, homes, forests), different land use around them, life in them, and because other streams flow into them. I plan on continuing this project in the future by building a computer model that might predict the water quality, just like you predict weather. This project is just a first step in doing so and a very important one. It showed to me that you need to be very site specific if you wish to predict water quality. Since each site is so different and there are so many factors that affect each one.

Abstract

The purpose of my project was to see whether two sites on the Chester River correlate or if their data moves together. I would be testing two sites, Gravel Run and Centerville Landing for a year, for each of six variables, air and water temperature, dissolved oxygen, nitrates, pH, and turbidity/clarity. I already knew some about these but researched them in greater depth. I also found out about the effects of salt water and fresh water when they meet. I did this because freshwater from Gravel Run runs into Centreville Landing where it and tidal waters mix, and it gave me a better picture of what the water quality should be there. After finishing my background reading I predicted that nitrates, pH, dissolved oxygen and the air and water temperature would all have a positive correlation of approximately 1. This is because the data should move together for each of these variables. The turbidity/clarity should have a correlation of –1 because they are testing opposites. After monitoring Gravel Run and Centerville Landing for a year I had all my data. When analyzing the data I found that the only two tests that had a correlation close to one were the air and water temperature. All the other tests did not correlate at all or, had a correlation so close to zero it did not matter. I found out from this experiment that each site is so different and they actually should not correlate at all. There are so many factors that affect water quality and these factors are very different for Centreville Landing and Gravel Run. For further study on this experiment I plan to take knowledge gained this year and put into a long term project in the future of making a computer model to predict water quality. By finding that two sites do not affect each other, I now know that I would have to make a model very site specific.

Comments