Estuary Net

A. Summary

1. The System

The Estuary project is concerned with monitoring estuaries, which is an important aspect of protecting and improving the environment.

Water quality of an estuary is determined by a complex interaction of chemical, biological and physical processes and includes factors such as temperature, salinity, dissolved oxygen, turbidity, and pH.  For example, respiration decreases dissolved oxygen, produces CO2, which in turn increases the acidity of the water (i.e., lowers pH). Decreasing the temperature of water increases solubility of oxygen.

This complex system is influenced by a number of external variables.  For example,

  1. the daily variation between day and night affecting air temperature, amount of light and heat energy that is available for the estuary.
  2. seasonal variations in these external variables
  3. the periodically varying tides
  4. specific events such as pollution, thunderstorms, sudden climate changes etc.
For more information see Introduction

2. Learning Goals

The major objective of Estuary Net is that high school students come to understand the scientific process and its contribution to solving real-world problems. During the project, students learn about variables involved in estuarine ecology, how these variables relate to one another, and how they affect water quality under ideal laboratory conditions.  Using this knowledge, the students, proceeding like scientists, develop hypotheses concerning relationships and time dependent variations in real estuaries. For instance, the students might predict a daily periodic variation of dissolved oxygen from their knowledge that the amount of photosynthesis in an estuary is influenced by the daily variation of available light.  These hypotheses are then tested and eventually modified by means of checking them with real data on estuaries.  Students can ideally use this same approach to solve authentic problems related to estuaries, including  environmental control and monitoring.
 

3. Available data

Data from 22 different estuaries are provided by the Centralized Data Management Office (CDMO) of the National Estuarine Research Reserve System (NERRS).  Variables are monitored at 30 minute intervals. Time series on all the system variables are available going back 2 to 3 years.  For some variables and time periodsdaily averages are available.

For more information about the data, see data and data archives.

4. Supports for data analysis

Software

The project offers no recommendations about particular analysis software for classrooms to use  but assume they will use whatever software they have available.

Subject matter knowledge

The project developer has summarized the relevant science of estuary ecology  in the document "estuarine ecology". The curriculum never refers to this paper, which is unfortunate since it includes information that would be helpful to teachers and students.

Data analytical knowledge and strategies

Most units have two major components.  The first part concentrates on theory and laboratory work; in the second, students test theoretical predictions by analyzing actual data from estuaries.   The second part is structured as a series of steps which offer students a general framework for analyzing real data in the context of scientific inquiry:
  1. formulate expectations
  2. download the data and graph the relationship
  3. interpret the graphs
  4. discuss the results
  5. check the results at different times and different locations
  6. create a report
Students probably find this structure helpful.

For more information about the structure of the curriculum and data analysis questions in the curriculum see data analysis in the curriculum.

However, we suspect that many students get bogged down in steps 2 and 3, as the curricula provide no specific suggestions about how to deal with various types of data and representations of data.   The graphs that students would need to explore their questions are complex and often superimpose information about seasonal variations, daily variations, and unique events like unusual weather patterns or introduction of pollutants. A student with little experience reasoning about data will likely be swamped by this flood of  information.

The graph below, loaded with information, is typical of what students might  produce:

Interpreting this graph at even a fairly rudimentary level would require students to the compare amplitudes and the wave lengths, look for superimposed trends and deviation from the normal trend, and determine whether there are dependencies among the three time series.  For them to make sense, students would have to interpret any patterns in light of both their knowledge about the subject matter and their results from the laboratory work. (For a detailed interpretation of this graph see our own analysis)

Subject matter questions to be answered by data analysis

The curriculum is divided into units.  In the first three units, students are introduced to 1) telecommunications 2) estuarine ecology, and 3) one of the sciences: chemistry, biology or earth science. Unit 4 is a summary unit, and involves the students creating a written report.   It is in unit 3 where students pursue various questions through analysis of data.   They first explore the relations among some of the measured variables in one of the selected sciences, as well the relations between the measured variables and either the tides or weather. Somewhat surprisingly, the impact of pollution is never in the foreground.   For the purposes of our own analysis, we chose the chemistry strand, which involved exploring the relations among dissolved oxygen, salinity, temperature and the tides.
 

Exemplary data analyses, expected results

The curricula does not provide any exemplary data analyses or indications of what students may find in the data.

5. Our exemplary data analysis

In our exemplary analysis, we chose to explore a chemistry activity at Level 1 of the curricula  (http://inlet.geol.sc.edu/level1.pdf). This involved using data available on the project's web site to study the relationship between temperature, dissolved oxygen, salinity, and depth. Below is a summary of our results.  For additional information about the analysis, the data used, and the results see exemplary data analysis.
  1. temperature, dissolved oxygen & salinity

    2. - The amount of dissolved oxygen (DO) varies during a day:  Due to photosynthesis, DO increases until the evening and then decreases during the night.
    - As mean daily water temperatures increase, daily means of DO decrease.
    - Average salinity and DO also vary inversely.
  2. salinity & depth

    3. Because sea water has a higher salinity than fresh water, salinity increases as the proportion of sea water in the estuary increases,
  3. dissolved oxygen & depth

    4. There seem to be higher levels of DO at greater depths.
  4. Temperature & depth

    5. The relation depends on the seasons, primarily because the temperature of sea water is less sensitive to temperature differences of the air than is river water. In winter, the sea acts as a heat reservoir, and the sea water is warmer than the water in the shallow rivers. Because high tide is produced by sea water, temperature of river water increases with the increasing proportion of sea water. In summer, water in the rivers is warmer than sea water because the former is more quickly heated by the sun. Thus in the summer, the temperature of water in rivers decreases with increasing proportion of sea water.
In our exploration of the data, we did uncover clear and interesting relationships among the variables.  This is good news, as the data do in principle permit students to perceive relationships that they would otherwise simply read about.  However, we spent considerably more than the recommended  2 class periods pursuing this question as there were a number of difficulties we had to overcome before we could detect these patterns:
  1. Daily variations are superimposed on seasonal variations and special events. Sometimes we had to transform the data or rescale the graphs to see the relations
  2. We had the 30 minute data and tabular data (daily means) for our analysis. It is probably very hard to see long term variations in the 30-Minute data. It is a disadvantage that the tabular data -- which contain daily summaries -- are not available anymore. Therefore students have to produce such daily summaries themselves as a first step, which is technically not so easily done with existing software. The patterns we discovered cannot be discovered in the raw 30-minutes data.On the other hand, the value of summarizing data for discovering structure can be experienced.
  3. The questions in the curriculum tend to be very open ended and provide no suggestions or hints about how to proceed.  The question we pursued was worded:  "Discuss the relationship of tides, salinity, depth and temperature."  We had to reformulate this into a series of  sub questions as it was not possible to consider the relationship among all four variables at a glance.
  4. There were no pointers to relevant subject matter to help in the formation of hypotheses.
  5. Many of the graphs we produced in exploring these question were not at all informative.  We can imagine students in the same situation getting quickly frustrated or confused.
Below are data analytical competencies we drew on in performing our analyses, which students would need in some measure to do something comparable.
  1. Seeing the estuary as a system of interacting variables, where relationships between 2 variables can be influenced (and disturbed) by other variables. Knowing how and when to aggregate and average data are important skills in dealing with such complexity.  Another strategy is to systematically check a relationship under various conditions
  2. Decomposing data into general trends and deviations from the trend, including perceiving and describing qualitative aspects of functions (increasing, decreasing etc.); seeing smooth functions in messy data, and being aware that such smooth functions are not uniquely determined by the data
  3. Manipulating  scatter plots by changing axes and scales, zooming, plotting several variables on one graph with well chosen (and perhaps different) scales, switching between point graphs and line graph depending on the data and the goals
  4. Attitude to deeply look into the patterns and details of scatterplots instead of being satisfied with superficial "quick looks"
  5. Dealing with time series with different periodicities in an elementary way; having available concepts such as amplitude and wave length with which to describe periodic characteristics and thus perceive subtle patterns in such graphs.
  6. Experiences with ideal periodic models such as trigonometric functions as base lines for comparison; without these ideal models, one may notice only  the cyclical pattern but not the more subtle, and critical, features of the periodicity; For example, knowing that standard sine functions have a constant difference between their relative maxima (constant "wave length") may help to recognize changing distances between relative maxima in real data
  7. Knowing that it may take considerable time in analysis before a clear pattern emerges and therefore not becoming frustrated or distracted by the many plots that in the end are not useful.

6. Summary from the perspective of data analysis

Estuary Net is potentially a suitable project for the high school.   Students have the opportunity of finding interesting and significant results from real data and in the process experience the nature of scientific reasoning.  However, we expect that without a knowledgeable and experienced teacher who can support students in the process of the data investigation, students would have little chance of making much progress with these data.  The project would be accessible to more classrooms if there were more entry level tasks included in the curriculum as well as exemplary analyses on the project web site to guide students and teachers.