Five of the original 9 sites continued in operation after 1997 for various periods. Two of these sites were operated continuously until September 2003. Three new sites were installed in the western part of Shark Valley in November 2001 for the purpose of testing regional model transferability. Additionally, an evaporation pan was installed at one site in April 2001 for comparing actual Et determined by the Bowen-ratio site with potential pan evaporation. All data collection ended in September 2003. The dataset contains the meteorological and evapotranspiration data. Additionally, tables listing model coefficients and goodness-of-fit statistics for site models for the period 1998-2003 are included, and tables listing a comparison for measured Et and Et estimated from the regional models.
Data is available by year for each of the collection sites.
The a_read_me file in the Data summary and data files for Everglades Et sites, 1996-2003 describes the format of data files of meteorological and evapotranspiration data. Additionally, tables listing model coefficients and goodness-of-fit statistics for site models for the period 1998-2003 are included, and tables listing a comparison for measured Et and Et estimated from the regional models.
This latest data release is different in format from the original release for all data from 1998 on. No changes were made in the 1996-97 data.
One change made in reporting format is that Et data from 1998 on are not smoothed by averaging over one or more measurement intervals. With this release data are provided at the measurement interval so that users may use whatever smoothing technique that is appropriate for the intended use.
Another change in format for data from 1998 on is that Et sums are provided for "raw" and "edited" 30-minute periods. The "raw" data refer to Et sums that have not been edited from computed results, although the Et sum may be an actual measurement that has passed all input-data screening tests (see WRI 00-4217), or may be a "gap-filled" value computed from the Priestley-Taylor site mode that was developed using only data that passed all screening tests. Data in the "edited" column have been edited graphically by comparing each value to the pattern of Et defined by the entire set of data during part of a day.
The final change in format for data from 1998 on is that a flag indicator is provided to show which 30-minute Et data are measured and which are model derived because the input data did not pass screening criteria.
The overall objective was to develop a process-oriented appraisal of evapotranspiration within the Everglades drainage unit, excluding agricultural and brackish environments. Specific objectives included: 1) Field measurement of evapotranspiration at a variety of sites encompassing a regionally representative range of environmental factors and 2) Verification and refinement of model using ET measurements at additional sites.
No changes were made in the 1996-97 data.
The data columns are identified by the text strings in the first row. The no-data indicator is -999. Time interval for all data except evapotranspiration (Et) is 15 minutes (site 1 and sites 4-9) or 30 minutes (sites 2-3.) For 1996-97 data, Et is given at 1-hour intervals (site 1 and sites 4-9) or 3-hour intervals (sites 2-3). For 1998 on, all Et data are given at 30-minute intervals.
A single file with the name "Ing_pan_daily.txt" is included with the Bowen-ratio site Et files. This file contains daily Et, in inches, from an automatic class A stainless steel evaporation pan at the Ing site.
In 2000, new sites were established in Shark Valley Slough, to test transferability of models developed using 1996-97 data, and to refine the understanding of factors related to Et. As of January 2003 there are five continuous Et sites in the Everglades National Park.
The original sites were selected to provide a network representative of the non-forested portion of the Everglades ecosystem in terms of plant communities, duration of water inundation (hydroperiod), and geographic coverage. Other factors in site selections were security and logistics. Sites in areas open to hunting and air boating were located in relatively remote locations and not on major air boat trails. Each site was located at the center of a circle of relatively uniform vegetative cover with a radius of at least 100 times the height of the upper air temperature/humidity sensor.
Stations were instrumented to provide data for: determination of total energy available for Et (latent heat flux) and convection (sensible heat flux); determination of the Bowen ratio (the ratio sensible heat flux/ latent heat flux), so that the amount of the total available energy that was utilized for Et could be determined; and characterization of meteorological conditions and Et-model development using ancillary data.
The array and arrangement of data sensors at the sites were dependent on whether the site was in open water or in dense, emergent vegetation. The major difference between open-water sites and vegetated sites is the method of determining the air-temperature and humidity differential with height, which is necessary for computation of the Bowen ratio. At the two open-water sites (sites 2 and 3), the air temperature and humidity differentials were measured from the water surface to a point 3-4 feet above the water surface. At the seven vegetated sites (sites 1, 4-9) the differentials were measured between two points in air, 3-5 feet apart.
At each site, sensor measurements were made automatically every 30-seconds and these measurements were averaged and stored onsite at 15- or 30- minute intervals. These data were then transmitted daily by cellular telephone to computer storage in the office. Data were reviewed on a daily basis to detect equipment breakdown and sensor malfunction. Site visits were made regularly for routine scheduled maintenance and cleaning, or more frequently when malfunctions occurred.
Data were collected from January 1996 through December 1997 for sites 1, 2, 3, and 5. Data were collected from January 1996 through December 1999 for site 4, from December 1996 through December 1998 for site 6, and from January 1996 through December 2003 for sites 7 and 8. Site 9, however, was installed in January 1997 to increase representation of drier parts of the Everglades; site 9 furnished data from January 1997 through December 1998. Only data that passed screening tests for accuracy were used to develop the models of Et. The screening tests were based on range limits, visual inspection of plotted net radiation, temperature and humidity readings to eliminate periods when sensors were obviously malfunctioning, and on criteria given by Ohmura (1982). These criteria specified that flux calculations are inappropriate if the calculated latent heat flux is in the opposite direction from the observed vapor-pressure vertical difference. Such a situation would indicate an error in determination of either the energy budget or the vapor-pressure or temperature vertical differences. Ohmura also recommended that Bowen-ratio calculations be rejected if temperature or vapor-pressure vertical differences are at or less than sensor resolution limits. Resolution limits for this study are 0.013 degree Celsius for vertical temperature differences and 0.003 kilopascal (kPa) for vapor-pressure differences. These screening criteria eliminated about one-half of the available data from model development, mostly because of sensor failure and resolution limits. Most of the data rejected because of resolution limits or flux directions were for night-time hours, when energy inputs, air-temperature vertical differences, and vapor-pressure vertical differences are all relatively low.
Sites were visited at 4-6 week intervals for inspection and maintenance. Maintenance generally included the following items:
*Ventilator fans - Clean and replace, if not operating. *Net radiometer domes - Clean and replace, if damaged. *Replace radiometer if water damaged. *Inspect radiation shields (air temperature and humidity) - Clean Air temperature and humidity sensors, Clean and replace sensors if necessary. *Water-level sensor - Raise float and check for proper response. *Rain gage - Check for obstructions, clear if necessary, test calibration periodically. *Water temperature sensors - Check for proper position and reading. *Net radiometers and pyranometer - Check for level, adjust if necessary. *Sensor exchange mechanism - Check for smooth operation, replace as necessary. *Field verification of air temperature, relative humidity, wind speed, and wind direction using handheld meters.
The net radiometer domes required the most frequent maintenance. These domes, made of soft transparent polyethylene, shield the sensors from moisture, wind, or debris that could affect sensor performance. Problems encountered included crushing by hail, pecking by birds, and gradual deterioration of the polyethylene. Domes were changed at 3-month intervals, or sooner if damage occurred. If the domes were cracked, punctured, or there was evidence of water penetration into the sensor, the entire net radiometer was replaced.
Air temperature and humidity sensors failed frequently during the first year of operation, due to corrosion of electrical contacts. A change in sensor design resulted in much-improved service life of these sensors during the second year of operation. The sensor exchange mechanisms were subject to occasional failure, generally due to mechanical wear or water penetration into the control circuitry.
Net radiation is measured directly by the net radiometers, but the measured value is affected by wind speed and must be corrected. The wind correction factor was calculated from wind measured at the sites using procedures described by C. Fritchen of REBS, Inc. in a personal communication. Soil heat flux was measured at all vegetated sites, but was not measured at the open-water sites because these sites were always covered by water, generally to a depth of more than 1 ft. At the vegetated sites the soil heat flux was determined from the sum of heat flux measured by a heat-flux plate buried 5 centimeters (cm) below the land surface and the change in heat stored in the soil profile above the plate. Water heat storage was calculated at all sites whenever water was standing on the water surface. At open-water sites with little or no emergent vegetation, the air-temperature and vapor-pressure differentials necessary for the Bowen-ratio determination are determined from measurements of water temperature at the water surface and air temperature and vapor pressure at a point 3 to 4 ft above the water surface. The water-surface temperature is measured by using a float -mounted thermocouple, and is assumed to represent the air temperature at the water-air interface. The vapor pressure at that point is assumed to be equivalent to 100 percent relative humidity. Because the differences between water surface and air are much greater than differences in the air over similar distances, the effect of air and vapor pressure sensor bias is negligible. Therefore, the sensor exchange mechanism is not required and only one air temperature /vapor pressure sensor is needed at such sites.
See WRIR 00-4217 (<http://fl.water.usgs.gov/PDF_files/wri00_4217_german.pdf>) for more detail and the formulas used in the calculations.
One change made in reporting format is that Et data from 1998 on are not smoothed by averaging over one or more measurement intervals. Although the Bowen-ratio method provides Et at 30-minute intervals, Et totals were previously (1996-97) smoothed by computing sums for 60-minute intervals at vegetated sites, and 180-minute intervals at open-water sites. This smoothing helps to remove "noise" from the Et data caused chiefly by imprecision in the determination of water heat storage and/or Bowen ratio during some measurement intervals. This imprecision is likely greater during periods of high water level, when small errors in estimating the temperature of the surface water column account for relatively large parts of the energy budget. Another factor is that the Bowen ratio is difficult to determine accurately during night-time or other periods when temperature and/or vapor pressure gradients are small. Because the Bowen ratio is the quotient of the air temperature and vapor pressure gradients, small errors in determining either of these gradients can cause relatively large variation in the computation of the Bowen ratio.
With this release data are provided at the measurement interval so that users may use whatever smoothing technique that is appropriate for the intended use. It should be emphasized that the 30-minute Et sums may at time indicate short-term variation that is probably not realistic. Averaging over several intervals is likely to produce a more realistic pattern of Et variation. Averaging Et at a daily interval has the advantage of removing most of the uncertainty in the water heat storage term, because water temperatures generally increase during the day and then decrease at night, with little net change in temperature (and water heat storage) for the entire day.
Another change in format for data from 1998 on is that Et sums are provided for "raw" and "edited" 30-minute periods. The "raw" data refer to Et sums that have not been edited from computed results, although the Et sum may be an actual measurement that has passed all input-data screening tests (see WRI 00-4217), or may be a "gap-filled" value computed from the Priestley-Taylor site mode that was developed using only data that passed all screening tests. Data in the "edited" column have been edited graphically by comparing each value to the pattern of Et defined by the entire set of data during part of a day. For example, if a spike in Et at night did not seem to be "real" in relation to available energy measurement and other Et values before and after the spike, the value was adjusted graphically by selecting the plotted value and "dragging" it into line with the pattern defined by the other data. This editing procedure is, of course, subjective, so the raw, unedited data are provided for those who prefer to use unedited data, or to use their own editing technique, such as data averaging or some other form of smoothing.
The final change in format for data from 1998 on is that a flag indicator is provided to show which 30-minute Et data are measured and which are model derived because the input data did not pass screening criteria.
The data collected at each site include the following:
Date_time - date and clock time of data collection (day/month/year/hour:minute) in Eastern Standard Time. The combined Date_time should import into most spreadsheet and data analysis applications as the appropriate date-time variable.
depth - average depth of water at the site, in feet above land surface. Negative values indicate depth of water below land surface. At some sites, depths were estimated by interpolation during periods of sensor malfunction.
rain - rainfall total, in inches
winds - average wind speed, in miles per hour
windd - average wind vector direction, in degrees clockwise from north
pyro - intensity of incomng solar radiation, in watts/square meter
nr - net solar radiation, in watts/square meter, measured with Q7.1 radiometers (REBs, Inc.). This sensor measures the difference between incoming (solar and atmospheric) and outgoing (reflected and longwave) radiaiton.
shf - average soil-heat flux, in watts/square meter
tair - average air temperature, in degrees Celsius
rh - average relative humidity, in percent
de - average gradient in vapor pressure, in kilopascals/meter. At all sites except the open-water sites (sites 2 & 3), the gradient is measured between two sensors separated by a vertical diaatance of about 1 meter, with the lower sensor about 1 meter or more above the top of the vegetative canopy. At open water sites (2 & 3) the gradient is measured from the water surface to a sensor positioned about 2 meters aabove the water surface. All de measurements are for 30-minute intervals, even though they are given at 15-minute intervals for most sites. The gradients are calculated by subtracting the lower-sensor measurement from the upper-sensor measurement, so that a positive gradient indicates increasing vapor pressure with height.
dt - average gradient in air temperature, in degrees Celsius/meter, These gradients are measured using temperature sensors co-located with the vapor-pressure sensors. All dt measurements are for 30-minute intervals. The gradients are calculated by subtracting the lower-sensor measurement from the upper-sensor measurement, so that a positive gradient indicates increasing air temperature with height.
wts - average water temperature, in degrees Celsius, at the water surface, measured with a float-mounted sensor. Water temperatures are not given when water level is below land surface.
wtm - average water temperature at a fixed position 1-2 feet above land surface (not recorded at most sites)
wtb - average water temperature at the water/land interface. Water temperature is not given when water level is belkow land surface.
Et_edit - cumulative evapotranspiration (Et), in inches for the reporting period. Some of the values have been edited to eliminate possible unrealistic values. The Et in each row is the total Et since thte last reporting period. A no-value indicator (-999) is listed for times within the Et-reporting period. To obtain an Et sum for a period using a spreadsheet or other software will require replacement of the no-value indicators with a 0 or a missing-value indicator.
Et_raw - identical to Et_edit except that no editing has been done; data values were measured if input data passed all screening criteria or are calculated for the Priestly_Taylor site model if the criteria are not met
Flag- type of Et value indicator; flag is 0 if the Et value is a measured value that passed all screening criteria and flag value is 1 if the Et value is from a site model rahter than an actual measurement
The data collected at the Ing pan site include:
Date - day/month/year Rain - total rain for the day, in inches Et_pan - total pan Et, in inches
U.S. Department of the Interior, U.S. Geological Survey
Comments and suggestions? Contact: Heather
Henkel - Webmaster
Generated by mp version 2.8.18 on Tue Dec 09 17:34:07 2008