Production of site imagemaps

The maps are based upon analysis of the water-level monitoring sites that are on the USGS Center for Water and Restoration Studies cooperative data-collection program that are continuously monitored. This set is further restricted to those sites that have had water-level data collected within the past 120 days. The basis for the retrieval is the second month prior to the date the maps are generated. For example, during April 2002, the target month was February 2002. For the sites, the following data are also retrieved from the project database:

1. County in which the monitoring site is located,
2. Water supply system monitored by the site (aquifer for wells),
3. The latest available 7-day averages of daily water levels at the sites for the target month, and
4. De-trended frequency statistics for the last week of the target month.

In addition to a large-area map showing the selected sites within south Florida, maps for each county, major aquifer, and the surface-water system are generated. The symbols for each site are determined by the type of monitoring system at the site, whether the site has recent data, and by comparing the 7-day running average of the stored daily values to the frequency statistics for the week of the year.

Basis of computations:

• Water levels are measured in feet above the National Geodetic Vertical Datum of 1929 (NGVD1929). This datum is not identical to mean sea level, although for many purposes the two are similar in usage.

• The 7-day average of water levels is calculated based on the daily maximum water level for the previous 7 days (at groundwater monitoring stations, daily mean gage height is used at surface-water stations). This statistical measure provides a more consistent presentation of water conditions for comparison to frequency statistics than using the most recent daily maximum water level. The most recent hourly measurement, or even an average of hourly measurements, would not be as useful because the frequency statistics are calculated on daily maximum or mean values.

• The frequency statistics for the site are obtained in a three-step process:

  1. The daily maximum (or mean, as appropriate) values collected during the past 25 years are retrieved, and a regression against time is performed to identify and calculate any existing long-term trends.

     The original choice of an early cutoff at 1975 reflects the limited extent of the ground- and surface-water networks in South Florida prior to the mid-1970s, and the change in USGS data storage methods circa 1974, making the data collected since then readily available in a computer-retrievable format. The project has retained the 25-year data window to provide a consistent statistical measure through time. Daily maximum water levels for ground water, and daily mean gage heights for surface water, are used because those are the data the USGS has reported from South Florida over the past several decades.

  2. The retrieved data are analyzed to determine if there is a long-term trend in the water-level record. If so, the trend is mathematically removed from the data. The data are then grouped by week of the year collected, and frequency statistics are obtained from the resulting data set using a standard statistical analysis package. The resulting data are then stored in the project database.

  3. Freqency statistics retrieved for comparison to the 7-day average are first recomputed to include the prevailing trend. While the frequency statistics are computed and stored as values referenced to the computed trend line, the water-level data themselves are stored as values referenced to NGVD1929. For example, even though at site X the measurements are 2 feet above the trend, the water levels may be 15 feet above NGVD1929. Obviously, the 2 feet and 15 feet values cannot be directly compared until the "de-trended" percentile values are recomputed to allow for the trend that was originally removed, making NGVD1929 its reference datum. Then the current 7-day average can be compared to the freqency statistics - the 10th, 30th, 70th, and 90th percentiles in the case of the maps.