Home Reports Start

2000 Annual Water Level Raw Data Report for Kansas

Prev Page--Executive Summary || Next Page--Statistical Quality Control

I: Acquisition Activity

Richard D. Miller and David R. Laflen, Exploration Services Section


A field crew from the KGS acquired data from 555 wells in 19 western Kansas counties during January 2000 (Figure 2). The measurement technique and overall responsibilities associated with the 2000 annual water level measurement program were completely consistent with previous years' acquisition activities of both the USGS and KGS (Miller, 1996; Miller, Davis, and Olea, 1997; Olea, 1997; Miller, Davis, and Olea, 1998; Miller and Davis, 1999). Data acquisition was broken into two field periods with the first (primary) extending from January 3 to January 9, the second running from January 13 to January 14. During primary acquisition, each well was visited and data collected (water level below ground surface, well condition, GPS, photograph, well characteristics, and general observational notes). Again in 2000, enhancement/replacement wells were located and measured during the first visit as well as some of the Quality Assurance (QA), Quality Control (QC), and missing (MIA) wells. The secondary acquisition was designed to finish the measurement of QA, QC, and other missing wells. Difficult-to-measure or initially inaccessible wells may have been visited multiple times. All raw data acquired by the KGS on the annual water level measurement program were compiled, digitally stored, and available for widespread circulation in their raw, unprocessed form (digital and/or analog) on or about March 1, 2000. This section represents a summary of all KGS activity related to the raw data, acquisition activities, and technique development.

Acquisition Logistics

The primary measurement trip was completed in 5.5 days by a crew of six people (with a seventh picking up problem wells and measuring new wells for the first five days). Each person was equipped with a vehicle, computer, GPS, cellular phone, field notes, steel measuring tape, and associated supplies. The far north-western counties (Cheyenne, Sherman, and Wallace) were completed the first day, requiring an overnight stay in Syracuse. The second day of measurements ended in Garden City and saw the completion of two counties (Hamilton and Kearny). The third field day ended in Liberal with three counties finished (Finney, Haskell, and Grant). The fourth day ended in Dodge City, with three more counties (Stevens, Seward, and Meade) completed. The crew next stayed overnight in Great Bend after completing three more counties on the fifth day (Gray, Ford, and Hodgeman). The sixth day was the last day of the primary trip and required only a half-day for four people to complete the last three counties (Ness, Pawnee, and Barton).

During the primary acquisition trip the six crew members were divided into three teams. Each team was responsible for 26 to 34 wells per day along predesignated closed routes. Each day's route was designed so team members would meet along the route after all the wells were visited. This method balanced workloads by compensating for different degrees of measurement difficulty (different lengths of time are necessary to measure individual wells), insured that no wells were overlooked during the primary measurement trip, and minimized the number of miles traveled per well and route. Distribution of crew members and routes was designed to insure that no crew member was more than 15 to 20 miles from another crew member. This increased safety and minimized the time necessary to complete a county. Once a team completed its route, the other teams were contacted to determine if assistance was necessary to complete all routes prior to sunset. After a team completed its route and fulfilled any requests for assistance, they proceeded to a predesignated motel. Approximately 11,500 miles were logged during the primary measurement trip with the average crew member visiting 16 wells per day and spending about 15 minutes at each well and about 25 minutes traveling from well to well.

QC/MIA/QA/Enhancement Wells

The QC/MIA/QA/enhancement well trip required two days for two people to complete. The QC trip was designed to remeasure wells classified as out of trend during the primary trip. The QC water level remeasurement covered 3 counties and included 3 wells. These 3 wells were selected based on the following criteria:

  1. calculated water level was up by more than 1 ft from historical and 2000 trend; or
  2. calculated water level was down by more than 4 ft from historical and 2000 trend.
Twenty-three wells identified as MIA were not measured during the primary trip for one reason or another. Of these, measurements were obtained from 16 wells. The remaining 7 represent the difference between all available to measure in 2000 (562) and the total measured (555).

The QA trip provided repeat measurement data for the purpose of statistical appraisal of measurement error. Fifty-six wells, or about 10% of all the wells measured by the KGS during 2000, were randomly selected by computer for remeasurement.

Again this year, an attempt was made to fill some of the spatially under-sampled areas within the network (Olea, 1999). To fully sample the High Plains Aquifer from water level data, 32 new well sites were needed. Input from the GMDs, DWR, and KGS identified 32 wells in areas that are spatially undersampled and that have sufficient historical and construction data to justify an attempted measurement. From those 32 candidate wells, 16 were sampled during 2000 as new annual network wells. The selection and measurement of these enhancement wells is critical to establishing a method for maintaining the health and continuity of the network through the insertion of new wells to replace unusable wells already in the network.

Well and Measurement Point Information

A few key well characteristics have been compiled for general information and incorporation into efforts to improve and maximize the network. Included when possible are statistics and characteristics for DWR wells as well.

Wells Measured in 2000
Type of WellKGSDWR
Unused (monitor or abandoned)103170
Depth to WaterKGSDWR
Less than 100 ft205440
100 to 200 ft232264
200 to 300 ft9547
More than 300 ft162
Drill Depth of WellKGSDWR
Less than 100 ft87245
100 to 200 ft84280
200 to 300 ft131144
300 to 400 ft9634
More than 400 ft8531
Measurement CharacteristicsKGSDWR
Oil on the water80 (14.4%)140 (17.3%)
Difficult measurement
(restrictions, snags, catches)
40 (6.7%)37 (4.6%)
Noted changes in restrictions42 (7.2%)7 (0.9%)

In-field confidence in a particular measurement was qualitatively determined through inspection of chalk cut while general accuracy was quantitatively appraised using historical and local trends.

Measurement Confidence
Wells Measured OnlyKGSDWR
Three times121
Four times70
Total MeasurementsKGSDWR
During primary and QC trips743845
During QA trips54N/A
Measurements JudgedKGSDWR

Network Continuity

An uninterrupted historical record is important for trend determination and analysis. Significant efforts were made to acquire measurements in all wells regardless of whether the well had been successfully measured in any of the last three years. Wells with problems that were not likely to improve between measurement years were identified by each measurer and then cross-checked by the QC measure. If both visitors indicate that a well needs to be removed from the network, it generally is dropped from the next year's list of annual measurement wells. A total of 15 wells were flagged by KGS field personnel during the 2000 acquisition season. Two of these 15 were replaced by personnel in the field at the time that they were classed for removal, thus leaving 13 wells from the 2000 season with failure flags.

13 Wells from KGS Portion Requiring Replacement and Reason
04S 41W 31 ACA 01Well silted in. No meaningful data. ('00)
06S 42W 08 CBB 01MP obstructed. Extremely difficult to measure. ('00)
19S 14W 23 BBD 01Well removed. Ground bulldozed. ('00)
20S 23W 32CDA 01Dangerous well with no real MP. ('00)
22S 24W DDD 01Household well. Confined space with electrical hazard. ('00)
24S 42W 04 AAD 01Dry. ('00)
25S 26W 30 ABC 01Plugged. ('00)
29S 29W 10 ABB 01Difficult MP. Several obstructions. ('00)
29S 30W 35 ACD 01Several restrictions. Extremely difficult. ('00)
30S 28W 33 AAA 01Vacuum sealed well. Impossible to measure. ('00)
30S 29W 23 CAD 01Several restrictions. Extremely difficult measure. ('00)
30S 31W 05 BBB 01Snags at water level. High risk well. ('00)
31S 34W 18 BBB 01Restrictions every 20'. High risk well. ('00)

22 Wells from DWR Portion Requiring Replacement and Reason
01S 26W 18DDB 01Difficult access.
02S 30W 23ADD 01Unable to measure for three years.
02S 36W 13DDD 01Unable to measure for three years.
04S 23W 26CCC 01Unable to measure.
07S 26W 12BAC 01Unable to measure.
11S 36W 06ADD 01Destroyed.
19S 37W 28ABB 01Unable to measure.
20S 10W 36ACD 01Plugged.
20S 35W 15BBB 01Unable to measure for three years.
21S 02W 28CBA 01Unable to measure for three years.
22S 09W 17BAB 01Steel plate welded to top of casing, unable to measure.
24S 03W 14BBB 01Dry.
25S 08W 19ADB 01Household well.
25S 09W 17BBC 01Unable to locate.
25S 11W 02ACB 01Household well.
28S 18W 05CDB 01No access.
28S 39W 33ACC 01Unable to measure.
29S 37W 03CDB 01Submersible pump.
29S 39W 24DDA 01Unable to measure for three years.
32S 42W 14CCC 01Difficult measure.
32S 42W 21BCC 01Obstructions.
33S 42W 05DCC 01Unable to measure for three years.

To maintain the longterm health and continuity of the network, both wells that should be replaced and acceptable replacements should be immediately identified. Field observations are the most important and insightful method of identifying wells needing replacement; however, statistical quality control analysis has provided valuable insight into the reliability of network wells measured in 2000. A group of 8 wells have measurements that appear erratic when compared to measurements in 1999. When the most erratic 3 of these wells are removed from the 2000 data set, all variation except aquifer code and use of weighted tape is eliminated. All 3 wells have at least 4 measurements. Repeated measurements suggest that these wells are very difficult to measure with confidence. Replacement of these wells is highly recommended. An additional 5 wells were identified as erratic, but were less extreme than the first 3. The statistical analysis turned up this relatively small number of wells that could have a major negative influence on analyses done using these data.

8 Wells from KGS Portion with Erratic Behavior in 2000
Most erratic:
23S 22W 07DAA 0124S 33W 18BDB 0227S 38W 15BBB 01
Less erratic, but still of note:
24S 23W 06AAB 0125S 36W 28CBC 0129S 34W 11ADD 01
24S 33W 18BDB 0227S 37W 04ABB 01 

For a third year the KGS crews were deployed with acquisition software running on notebook PCs interfaced to GPS units. This system has been under development at the KGS for the last four years. It is designed to provide the measurer with historical data, warning messages in the event of an out-of-trend measurement or incorrect well location, automated depth-to-water calculations based on hold and cut, and real-time tracking and vehicle location displays. The system's primary focus is on insuring that the correct wells are measured and on enforcing complete well site documentation. The GPS units provide the associated computer with a tracking log, which permits time and location of each measurer to be determined throughout the day. During the 2000 measurement program the KGS crews logged over 11,000 miles in 6 days (Figure 3). GPS-measured latitudes and longitudes can be extracted from this log and used to improve the accuracy of well locations in WIZARD.

Figure 3--Example tracking logs of all measurers as recorded by the GPS/computer systems located in each vehicle (not to scale).

map of Kansas showing vehicle movements


Wells are grouped by county and then cataloged according to well ID (township, range, section system). All measurements taken during the annual measurement period are reported in Appendices C and D. In some cases a single well may have as many as four recorded measurements. The best value was determined by the field person(s) who measured the well. The best measurement was based on quality of cut, difficulty reaching the hold line, ease of retrieving the tape from below water level, pre-cut moisture, level of confidence that the tape was hanging unimpaired in the borehole, and accuracy of measurement point hold.

The raw data tabulated in this report are organized into five appendices. A brief discussion of the contents of each appendix follows.

Appendix A Contains a summary of information for all wells measured by KGS. This appendix includes only county, well legal description (ID), and 2000 depth of water below ground surface (BGS). QA measurements (QA) are identified.
Appendix B The same information as Appendix A, but for DWR wells.
Appendix C Contains all measurements and characteristics taken at every well in the KGS portion of the network. Included are the following categories of information: County, legal description (ID), GPS Lat, GPS Long, hold point, cut line, measurement point (MP) elevation, 2000 depth to water BGS, initials of measurer, measuring agency, and all comments.
Appendix D The same information as Appendix C, but for DWR wells.
Appendix E List by county of wells (legal description, ID) measured as part of the QA program. The list includes the primary measurement of DTW and the QA measurement of DTW.
Appendix F Contains the same information as Appendices C and D for the enhancement wells.

A direct comparison of primary measurements and the QA measurements made by the KGS reveals important information about the accuracy and repeatability of information in the database (Appendix E). In general, the time separation between the primary and QA measurements is one to seven days. This remeasure information, along with parts of the primary data, are integral to quality control discussions documented in subsequent sections of this report.

KGS Data Acquisition Summary

In summary, this year's effort by the KGS staff to acquire annual water level measurements met or exceeded most expectations. Based on preliminary analysis:

1) Systematic errors are an important indicator of erratic wells that should be removed from the network to insure that the network is accurate and that confidence can be placed in the data (see Statistical Quality Control Measurements section of this report).

2) More data have been acquired in each of the last four years (1997-2000) than in the previous 7 years (1990-1996).

3) The availability of digital and analog measurement data (for entire network) to DWR, KGS Geohydrology staff, and GMDs has been dramatically improved over years prior to 1997. These data will be provided on CD-ROM.

4) Field acquisition time has been reduced (2000 required 5.5 days; 1999 required 5.5 days; 1998 required 6.5 days; 1997 required 8 days; the historical average of time required for field data collection has been about 2 months).

5) Long-term improvements to the network and database are being made.

  1. All wells measured in 2000 have GPS lat and long and photographs;
  2. Errors, missing information, and incorrect information in both KGS and USGS historical database are being identified, evaluated, and corrected;
  3. Spatial distributions of wells based on aquifers are being evaluated and re-mediation efforts are underway for the first time;
  4. Wells are being added by the KGS to the network to fill "holes" as determined by spatial analysis;
  5. Acquisition techniques and procedures are being modified and improved based on statistical analyses;

f) A Kansas water well database (WIZARD) is being established for quick access to all water level information in the state through a Web site; and

g) The Quality Control program is providing valuable information about measurable wells that have erratic behavior and which should be removed from the network.

6) The total program costs for 1999 were about 75% of 1998 and 1997, which were both consistent with the amount paid to the USGS in 1996 to complete this study. When calculated using an equivalent* product, the cost to the Kansas Survey for the 1997 and 1998 water level data is about half that assessed by the USGS in 1996 and less than one-third of the USGS 1996 costs for the 1999 acquisition season. (*Equivalent product cost does not include extra program activities such as QA or QC data acquisition, analysis, or computer and GPS equipment.)

Development of a Kansas water well database was undertaken by the KGS in an attempt to make information quickly and easily accessible to the general public about water wells, both those that are part of the annual network and many that are not. This database will hopefully provide quick and accurate information about potential replacement or enhancement candidates throughout the State. At the present time this database is still under development, but should be fully operational by next field season. The database, by design, will include all significant information contained in the USGS's GWSI database, the KGS's KIWI database, and the KGS's WaterWitch database. It is the intent of the database's designers to incorporate portions of DWR's WRIS, KDHE's WWC5, City of Wichita, and each of the five GMDs' water well databases. Once this database, named WIZARD, is complete it should contain the most inclusive listing anywhere of information on water wells in Kansas. It is the intent that frequent uploads from each of the parent databases will insure that information in the database is current.

Prev Page--Executive Summary || Next Page--Statistical Quality Control

Kansas Geological Survey, Water Level CD-ROM
Send comments and/or suggestions to webadmin@kgs.ku.edu
Updated March 9, 2000
Available online at URL = http://www.kgs.ku.edu/Magellan/WaterLevels/CD/Reports/OFR0010/rep02.htm