Explanation of the Groundwater Database (GWDB)
The Texas Water Development Board (TWDB) Groundwater Database (GWDB) represents many years of collection efforts. As of March, 2013, it contains information for nearly 140,000 sites and includes data on water wells, springs, oil/gas tests (that were originally intended to be or were converted to water wells), water levels, and water quality. The purpose of the TWDB's data collection efforts over the years has been to gain representative information about aquifers in the state in order to support water planning from the local to a more regional perspective. Although the nomenclature for a water well identification number, or "state well number," may have promoted the impression that all water wells ever drilled in the state should be assigned such an identification number (and, by extension, reside in the TWDB's database), TWDB employees responsible for the creation of the database did not intend that such a capture of information occur or believe it desirable. The wells in the database represent less than 10 percent of the wells that actually exist in Texas.
TWDB estimates that between 1.35 and 1.5 million wells have been drilled in Texas since 1900 (Pie chart showing well records in Texas). Approximately 800,000 water wells, drilled between 1962 and 2002, have scanned driller's logs and are accessible in the Texas Commission on Environmental Quality's Water Well Report Viewer. However, these wells have not been located precisely or accurately because many are denoted only with a hand-drawn sketch or on a small county road map, if the location is indicated at all. No more images are being added to this viewer.
The Submitted Driller's Reports (SDR) Database, in existence since 2002, contains the drill reports (and logs) that drillers are required, by the State, to file upon completion of water well drilling since drillers began using the application in 2002. This database continues to experience fast growth as the percentage of drillers registering online increases. Currently, TWDB is adding records of wells and springs (whether newly drilled or existing) to the GWDB at a slower rate than occurs in the SDR due to a number of different reasons. Drilling is a commercial activity that has not abated in the past fifty years in Texas. Drillers are required to file information about wells they drill that they believe to be correct with no additional oversight, whereas data entered in the GWDB are subjected to more rigorous Quality Assurance/Quality Control by a limited number of State employees and/or staff in cooperating local groundwater districts and the US Geological Survey.
Developing the GWDB in the mid-1980s required the review of all reports and paper files to ensure that all pertinent site data for water wells, springs, and oil tests were entered in the database. However, data from many primary sources of information, particularly old Bulletins and Miscellaneous reports, have not been entered. The table that follows this explanation lists which counties have had data entered from older reports (and at what percent, if appropriate), and which have not. It also lists which counties have had water quality data reviewed.
To download the most recent copy of the GWDB or to view GWDB reports, click here.
The information in the GWDB has a variable range of accuracy as data collection methods and data maintenance have changed over the years. Knowledge of this information can help ensure appropriate interpretation and application of the data. Data inaccuracies that might exist are constantly being corrected, as staff time allows, in order to provide the highest possible quality data to users.
Please take a moment to review this explanation that describes some of the possible idiosyncrasies associated with specific database fields.
Well Record Data Fields
State Well Number (SWN):
Currently, all wells and springs must have this identification number to be included in the TWDB GWDB. The term is misleading because, as explained, not all wells drilled in the state have been assigned or were ever intended to be assigned such an identification number; furthermore, the TWDB has also used this identification for spring sites.
In general, the TWDB strives to maintain a database that is not merely a register of well locations (with depth and elevation), although much of this database can be defined as a legacy register, but also one in which wells have accurate water level, water quality, and, when possible, accurate completion data.
Site Location - Coordinate Accuracy:
Wells with a coordinate accuracy of '0' have been located using Geographic Positioning System (GPS) instruments. Close to 100 percent of the wells that are currently in the water level observation network have this accuracy, and all wells in the water level network or sampled for water quality within the past ten years have coordinate accuracies of '0' or '1' (+/- 1 second).
Wells with a coordinate accuracy of '5' have the least accurate latitude-longitude data. Latitude and longitude were assigned to locate the well in the center of a 2 1/2-minute grid on a topographic map. The Data Team Lead has worked to upgrade the majority of these coordinate accuracy codes, and currently code '5' sites account for 752 of the nearly 139,000 sites.
Codes '2' - '4' have been assigned to wells for a number of reasons, such as fact that latitude and longitude were entered prior to the addition of the coordinate accuracy field, data are from an outside source and their coordinate accuracy was not captured, and locational data were digitized from topographic maps by interns/technicians and the accuracy is unknown.
Public Water Supply Wells:
Of the approximately 22,000 public water supply (PWS) wells currently in existence in Texas as of March, 2013 and as tallied by the Texas Commission on Environmental Quality (TCEQ), TWDB has nearly 13,000 PWS wells in its groundwater database. They have been assigned PWS identification numbers by the TCEQ and its predecessors, but not all are present in the Remarks field of the TWDB database. The water level and water quality data associated with these wells that have been primarily collected through TWDB programs may have the same inaccuracies described in the following sections. However, the site location information and the current status of these public supply wells is ultimately the responsibility of the TCEQ. Any of this type of information should be cross-checked with the more up-to-date information through the TCEQ's Water Utility Database.
Other well record data fields:
Owner: Current only at time of inventory.
Aquifer: Most aquifer picks are correct; however, aquifer codes in some areas are in need of refinement. Many of these codes were assigned prior to a redefinition of aquifer names.
Date Drilled: Usually accurate with some exceptions (owner/operator memories are not reliable). A driller's report is most accurate. The year with day and month can also be variable. Drilling for many large capacity wells may have started on one day and not finished for many days or months. Therefore some water quality analyses or water levels may have a date earlier than the well's completion date. This is not an error.
Depth: Accuracy depends on the source. Problems with this field commonly start with the depth "interpretation" (memory) of the person who originally reported the data to an agency employee inventorying information at the site. Also, depth can change after initial inventory due to deepening or plugging back, but might not be apparent during subsequent visits. This is not uncommon. Water levels that are deeper than the original well depth or large changes in water levels or water quality over time are indicators of depth changes. However, the TWDB is attempting to reconcile these discrepancies. Nearly 5,000 seeming depth to water-total well depth and date drilled-data collected discrepancies have been reviewed and/or reconciled since 2010.
Completion Intervals: These data, if they exist, need a considerable amount of work. Some intervals were put into the Remarks field because this is how they were originally reported in groundwater reports.
Land Elevation: Check coordinate accuracy to determine the accuracy of land elevation. Numerous miscellaneous water level measurements are likely to contain suspect altitudes. Do not hesitate to question anomalous data points.
Well Type: This field should not be confused with the Water Use field, although because some of the same code letters are used for each field and the definitions could be considered identical, it's easy to confuse the two categories.
Pump information, water use and other available data: The data are current only as of the date inventoried. Much of these data were not in the original records of wells when entered in the database. Also, some of these fields may have changed over time. Many large capacity wells completed before 1950 are no longer in use; however, no one has revisited them in recent times to verify their status. The date collected and date updated fields are blank in many cases and require research of scanned images of the original well schedules-inventory records to determine date of the original site visit.
Water Quality Data
The following is to make users aware of some of the many idiosyncrasies of the water-quality data, primarily before 1989. The old WD-5100 water-quality data storage system on the Sperry UNIVAC began in the 1960s and evolved over time. It was first converted to an Informix system in 1988 and subsequently migrated to SQL server. During this time, several changes have taken place in terms of data issues and ranging from decimal places to system performance. As with the basic well data, updating and correcting any database deficiencies is an ongoing process. Please note certain discrepancies associated with the following fields, typically stemming from database structure demands.
General: The field for the less than (<) symbol was added in 1988.
Date: If the day of the month is not known, 00 appears in the day field; this allows this field to be queried and the data accessible.
Sampled Interval: Used only if samples were collected from intervals other than the completed interval.
"B" or "U" Analysis: Balanced analyses are ideal; however, many partial data, even in "unbalanced" analyses, are correct. Also, if any one of the following fields is blank, the entire analysis will be considered "U": Ca, Mg, Na, CO3, HCO3, SO4, Cl, F and NO3. The most common fields left blank are the carbonate/bicarbonate values. Furthermore, some me/L values in older analyses performed by TDH were hand calculated and incorrect.
The accuracy of the analyses may be checked because the solution must be electrically charged: the sum of the anions in me/L should equal the sum of the cations in me/L. The relationship is usually expressed as a percentage, where
Balance = (cations - anions) / (cations + anions) x 100
If the balance is < 5%, the analysis is considered balanced. If the analysis is greater than 5%, a "U" will appear in the database. Reasons for being unbalanced include:
- The analysis is not good.
- Major constituents were not used in the calculation.
- The water has a very low pH and the hydrogen ion was not included.
- A significant quantity of organic ions were present.
- Data entry/calculation errors.
Silica: This compound is inert and contributes to TDS.
Calcium/Magnesium: Contributes to hardness. Used in sodium absorption ratio (SAR), relative sodium content (RSC), and percent sodium (%Na) calculations. Round-off "error" can be corrected using original chemical analysis report for the site as long as milliequivalents per liter (me/L) values are known. Less than (<) symbol needs to be used in many cases. If Ca/Mg values are corrected, the SAR (Sodium Adsorption Ratio) value is no longer valid.
Sodium: Many of the older analyses had calculated sodium/potassium values, which required that their analyses be balanced. A small "c" in the flag field in front of the sodium field designates that this is a calculated value, not a measured one. However, not all calculated sodium values are identified as such; to identify that this parameter was calculated, the scanned image of the original analysis report must be researched, if available.
Potassium: The original form submitted to be keypunched had one decimal place for potassium. In the 1986-7 time frame, a new Texas Department of Health (TDH) printout form was overlaid on TWDB forms when mailed to the agency. On this overlay, the potassium appeared as a whole number. Keypunch folks typed in the whole number without one decimal place; so the value, for example, would show-up as 0.50 instead of 5.00. This problem was fixed, but for some analyses taken during the 1986-7 period, it may be continue to be incorrect.
Strontium: This constituent contributes to the hardness value.
Carbonate/Bicarbonate: There was no place for carbonate in the original databases. Analyses in the database without a CO3 value are essentially unbalanced. Many reports with analyses in them do not have a CO3 field. Generally, if the pH is below 8.35, the CO3 value = 0.00. Entering analyses from outside cooperators has contributed to the problem if the pH value and/or CO3 values are unknown. In calculating the sum of constituents, the HCO3 value is converted to CO3 (HCO3 x .4917) and added to the remaining major anions/cations. Both HCO3 and CO3 are calculated from phenol and total alkalinity. During data entry, if the phenol and total alkalinity are known, the CO3/HCO3 fields are blank. An extremely high or low pH will result in H or OH ion concentrations, and these values should be found in the infrequent constituent file.
Sulfate/Chloride: Some less than (<) values are absent that should be present. Original data entry allowed entry of only whole numbers; in the late eighties the database structure allowed for entry of numbers with two decimal places. Older values may have to be upgraded when found. For the most part, however, these constituent values are correct.
Fluoride: Constituent values have two decimal places. This value is only important should the value be greater than 2 mg/L.
Nitrate: Sample collection techniques have changed dramatically for this constituent. It is often hard to compare historical with more current data. Furthermore, values for both nitrate as N and nitrate as NO3 exist in the database. Contracted labs conducting analysis for the TWDB report nitrate as N and as NO3. Nitrate (N) is in the infrequent constituent table, and the value for nitrate as NO3 is in the "waterqua" table. TDS is calculated using nitrate as NO3. Most nitrates in the database cannot be converted back to nitrate (N) unless the collection techniques and lab procedures are known, and these may or may not be verifiable on scanned image documents of lab reports.
Please note the different primary drinking water standard associated with N vs NO3. Nitrate has a primary drinking water standard of 10 milligrams/liter as N but 44.27 milligrams/liter as NO3.
Derivation of conversion factor:
Atomic wt of N = 14.007
Atomic wt of O = 15.999 x 3 = 47.997
Sum = 62.004 / 14.007 = 4.427
nitrate (N) x 4.427 = nitrate (NO3).
Dissolved Solids: (sum of constituents) This is calculated based on the values, in mg/L, of the major anions and cations, silica, and 0.4917 of the bicarbonate. Nothing is added into the "TDS" from the infrequent table. However, some high values that might be considered as contributing to the TDS, while not included in the TWDB's formula, are Fe, Br, B, Ba, and Zn. If a sample is missing one or more major anions or cations so that the analysis is unbalanced, a TDS determined by residue can be entered into the dissolved solids field. However, if all constituents are present, the TDS is calculated and replaces anything else in the field.
Alkalinity: Carbonate and bicarbonate are determined from this measurement. In most cases, this value is determined by the lab. Field alkalinity, found in the infrequent table, has been determined in the field.
Hardness: Hardness is determined (/calculated in the database) primarily from Ca, Mg, and Sr values. Barium, present in the infrequent table in micrograms/liter (ug/L), generally constitutes a much smaller percentage of the total cations contributing to hardness. Although not as abundant as strontium, in some areas it is fairly high and does contribute to hardness.
The hardness formula used in the database is as follows:
3 me/L (Ca + Mg + Sr + Ba) x 50.05 = hardness
3 me/L (Ca + Mg) x 50 = hardness (where Sr and Ba have not been determined)
(hardness - alkalinity) x 50 = noncarbonate hardness
Percent Sodium: This is a calculated value. The formula is as follows:
Na x 100 = % Na Values are in me/L
Na + K + Ca + Mg
Soil Absorption Ratio (SAR): This is a calculated field. The formula is as follows:
(Na/23) / √ (Ca/40+ Mg/24)
Specific Conductance: Analyses run by the Texas Department of Health (TDH) for the TWDB may be inaccurate. (This lab was used by the TWDB in the 1980s and early 1990s.) When the analytical results were returned from the TDH lab, in many instances the specific conductance values were less than TDS, which is incorrect. Instead, the diluted conductance, as eventually corroborated by the TDH, was the more accurate value. TWDB attempted to switch all the conductivity values in the database, but was not entirely successful, and a few incorrect values still exist.
TDS (mg/L) approximately equals (conductivity x A), where A = 0.46 to 0.76. Waters high in sulfate can be as high as 0.96. The value of A can be determined simply by dividing TDS by conductivity for sampled sites in the surrounding area that have similar hydrologic attributes.
pH: Values of pH in this field can be either lab or field determined. Since 1989, TWDB has measured pH in the field.
Fe, Mn, B: These are the only trace metals that were allowed to be entered in the database during the 1980s. The main problems associated with these values are: 1) they were stored in mg/L with one decimal place; 2) even when TDH measured these parameters with greater precision, TWDB forms could not accommodate additional decimal places; 3) no "less than" (below detection limit) flags existed; 4) round-off errors occurred; 5) conversion from milligram to microgram per liter compounded the round-off errors; and 6) there was no differentiation between total or dissolved concentrations. These values may not have been corrected - check the water quality review table. However, their original values may be located in scanned images of lab analyses.
1. Although an original TDH analysis for Mn may have been reported as <0.05 (mg/L), in the TWDB database this would appear as 0.1 (mg/L). There were no < flags, and only one decimal place existed for data entry, thus the value was rounded off to 0.1 mg/L. In conversion to micrograms per liter, the value became 100 ug/L. Secondary drinking water standards for manganese are at 50 ug/L. So in these cases, the actual manganese value that was below standards at < .05 mg/L, now (incorrectly) appears as above standard at 100 ug/L.
2. A boron value of 0.24 was entered as 0.2 mg/L. In converting the value to ug/L, it became 200. We now have to go back and make the value 240 ug/L. This type of situation also applies to iron.
In summary, when using the TWDB groundwater quality data, always check reliability, lab and collector codes. In addition, please note:
- specific conductance values from analyses with lab ('01', or the TDH lab) and collector ('01', or the TWDB) have been changed to the diluted value
- mg/L vs ug/L
- difference between nitrate as N(nitrogen) and nitrate as NO3(nitrate)
- Fe and Mn values (especially of 0.0, 20.0, 50.0, or 100.0) are suspect, due to the errors mentioned previously
- NO3 values of 0.4, 0.1, or 0.04, that may be missing a "less than" (<) flag
- unbalanced analyses, whose original analyses should be checked
- less than (<) symbols were not present in earlier database versions
- most older analyses had the sodium (and potassium) calculated, so these analyses may seem balanced but may not be
Historical Data Entered and Water Quality Reviewed by County
This table lists which counties have had historical data entered and which have had thorough water quality reviews:
Water Quality (WQ) Review - YES in this field indicates all analyses have been checked and upgraded in including the addition of metadata such as collector, reliability, and lab codes, with remarks added where pertinent. When available, Fe, Mn, B, and conductivity values were corrected. NO in this field indicates that certain constituents are wrong and that most metadata codes will be blank. Note potential errors discussed above.
Historical (History) Data Entered - Old TWDB publications (mainly the Miscellaneous, or M, and Bulletin, or B series) contain much water level and water quality data that were lost due to their attachment to sites that were never assigned a SWN (as they are configured today). Thousands of water quality analyses and water-level measurements from the 1930s through the 1960s have not been entered in the database. This table lists which counties have had historical data entered. YES (always with report source(s) listed) indicates the data have been reviewed and entered; a percent indicates what percent (with report source(s) listed) has been reviewed and entered; NO indicates that no data have been reviewed and entered, although the source(s) to be reviewed are indicated; and NONE indicates that no source documents with old data exist.
|61||Cameron||NO||SOME OF B5403|
Please contact the Groundwater Data Team with any groundwater data related questions.