IAN Science Communication Forum

Please login or register.

Login with username, password and session length


Author Topic: Tampa Bay Qs NOAA As about completing survey  (Read 5296 times)

0 Members and 1 Guest are viewing this topic.

Offline suzanne

  • Member
  • Posts: 6
Tampa Bay Qs NOAA As about completing survey
« on: November 23, 2005, 11:55:40 am »
These are comments and questions from the Tampa Bay group re: filling out the NOAA Eutro Survey, answers from Suzanne Bricker (NOAA) and then additional Response from the Tampa Bay team if deemed necessary.

Tampa Bay Team: Ten scientists and managers working in Tampa Bay collectively filled out the Eutro Survey today.  We have several qualifiers, comments and questions.  All of these are from extended discussion among our group.

NOAA: Wonderful! This is the type of collaboration I hope that others will also engage in. Thank you very much for your support of this effort.

For the water quality conditions (chl a, DO, etc)
1.  We decided to use the salinity measured at the time of the sample, rather than the map, to put the sample into the tidal fresh, mixed or seawater bins. We used the maps to define the areas for macroalgae cover and seagrass, though.  Our stratified random WQ monitoring design does not lend itself to revisited sample sites.

A. This is fine. Do you also sample annually since I also should make it clear that if there are only samples taken from the NCA index period then probably a mean value is more representative since the 90th and 10th percentiles will lead to biased values.

Response: The Tampa Bay monitoring program samples monthly throughout the year.

2.  We found that the 90th percentile instructions were vague, so we looked at the data from several different angles, and have used the following: for each parameter (i.e. chl-a and DO) we found the respective percentile (90th and 10th, respectively) for each year (1999 - 2004) from all instantaneous observations taken for that particular year. This gave six, unique percentile values. From these six yearly percentile values, we calculated the mean, median, and most extreme value. As a group, we agreed that the median of the six, unique yearly percentile values would be the most appropriate value to enter into the survey because it would describe the median percentile conditions in the estuary over the 1999-2004 time period.

It would probably be very important to let the other survey participants know that you are not looking for average annual chl a, but for the 90th percentile?in other words, only 10% of the sample points are WORSE than what you want in that slot.  Right, and we are assuming that these values are the anomously high ones and not representative of ?typical? bloom concentrations. Also, we are very fortunate to have a water quality program with scientists who are versed in statistics and SAS, so were able to do these analyses?I?m not sure that all areas will be as fortunate. The 90th percentile isn?t that difficult to determine if you sort the samples (well, depends on how many samples you have) low to high and then take the 90th percentile value?.or by station it?s the same thing which is the best way to determine the spatial coverage but only in the sampling stations are stationary?.in your case this probably wouldn?t work.

3.  The spatial coverage questions are quite laborious to answer also-  it takes a pretty sophisticated GIS program to do so.  Lat/longs are needed, as well as some sort of areal estimator.  We?re still working on this for our survey.

A. For the spatial coverage, we?ve also used (if we did not have access to adequate GIS programming) a percentage of systems that show the higher levels (if the sampling locations are equally spaced) to estimate the spatial coverage or even heuristically by looking at the distribution of samples on a map depiction of the salinity zones. There are several ways to do this and I do not want the lack of GIS polygonal analysis capability to preclude a response. Especially considering there are different grades of certainty that can be used to indicate the use of a less robust method.

4.  For the Frequency of Occurrence, we considered it ?periodic? if three or more of the annual average for the six years were worse than the median 90th percentile value- again, no direction for how to call this that we could find.  Episodic was 1 or 2 years of the 6 in our call.

A. The frequency of occurrence is to determine whether you have blooms that reach a particular concentration every year, which in some Atlantic coast systems is a good bet that the winter spring bloom for instance will be of similar Chl a concentration each year, but only at a particular time each year. This in contrast to episodic where you get high concentrations/blooms only when the wind blows from a particular direction for 3 days or something like that?.and persistent is pretty clear, think of Tijuana estuary with the Mexican inputs being rather high in nutrient concentration at all times.

Response:  For DO, we have top, mid and bottom measurements, but have used the bottom measurement following the survey directions. And again, the survey asks for the level for which 90% of the data points are better.  As for the other WQ parameters, we used the median of the six, unique yearly percentile values.

5.  We didn?t know how to answer the question ?Has the overall eutrophic conditions of the estuary impacted any living resources??.  We have seen an improvement in living resources (seagrass acreage, chl a concentration) since 1999, but are still only about half-way to our goal of restoring seagrass to 1950s levels.  So what time frame are you looking for here when you ask ?how much of an impact has eutrophication had on the living resources???  If it is the last 5 years, then it would be ?no impact? (we?ve seen increased living resource indicators), but if we look at the long-term, our bay?s living resources are still heavily impacted from eutrophic conditions in the past.  We?ve put ?unknown? in there for now, until we can get clarification from you. 

A. For this question, we mean for the (approximate) decade from the early 1990s to the early 2000s. How would that mesh with the most recent 5 years? And then there?s always the comment area where you could specify?.as well as the discussion at the workshop. One chapter of the report will be dedicated to case studies and (as you know) I?ve always thought that Tampa was a great story?.can we use this in this report too?!