Agricultural Irrigation Water Demand Forecast

Procedures for measuring irrigated acreage for the baseline year and mapping of individual irrigation fields and water sources.


James E. Hook, Co-PI, Shane Conger, GIS Specialist, and Annie Horak, Technical Assistant,
The University of Georgia, NESPAL.
Tifton Campus

Acknowledgement: The UGA Ag Water Demand team would like to thank the people who provided their assembled data and provided feedback and assistance to assure the most complete coverage possible for Georgia's irrigated areas. These include David Eigenberg, Luke Crosson, and John Segars of the Georgia Soil and Water Conservation Commission (SWCC), Cliff Lewis and Danna Betts of EPD's Agricultural Permitting Unit (APU), and Mark Masters and Marty McKimmey of Albany State University (ASU) and the Flint River Water Planning and Policy Center (FRWPPC).

Baseline Irrigated Area:

Three sources of data were used to create a map of each county's irrigated areas, and, in many cases allowed identification of type of irrigation and type of crop. These maps formed the base acreage from which future year's irrigated area projections were made for computing the Ag Water Demand. Map sources included the Georgia Department of Natural Resources Environmental Protection Division's Agricultural Permitting Unit, Georgia Soil and Water Conservation Commission's Agricultural Water Conservation and Metering Program, and Albany State University's Flint River Water Planning and Policy Center. These were supplemented by new mapping by The University of Georgia's National Environmentally Sound Production Agricultural Laboratory. Through their cooperation and support we were able to compile a statewide inventory and map of Georgia's irrigated land.
Irrigation Development in Georgia
A background in irrigation system types and
timeline of irrigation growth in the state.

EPD-led irrigation mapping
Well and Pump Locations:

Georgia's Environmental Protection Division - Agricultural Permitting Unit (APU) provided two sets of GIS data. The first contained the locations of all permitted water withdrawal sources. Unlike municipal and industrial withdrawals, permits are required for individual wells and, in most cases, individual pond and stream pumps. Pumps and wells may be used together to supply individual or interconnected irrigation systems. These source locations, required for each permit enabled us to identify water sources for many of the previously unmapped irrigation areas. At the start of this effort in November, 2008, APU records contained 15,582 wells and 16,736 surface pumps. For 26,731 of these, a geographic location was available. Those include almost all most permitted and active withdrawal locations, and the majority of applications for new withdrawals.

Irrigated Fields:

In addition to well and pump location, EPD provided us their GIS coverage of fields mapped in past efforts to identify irrigation land. Sample EPD mapped fields are shown in yellow in images to the right. EDP has conducted field or GIS based mapping to gather information on actual irrigated acres. Agricultural Water Withdrawal Permits list the irrigated acreage that can be used with any permitted water source. However, most farmers use only part of that acreage. Additionally, some permitted water sources become inactive and unused. The permits remain in EPD records because they were issued in perpetuity. At any point in time, actual irrigated acres will cover 25 to 35 % less land than permitted acres. To aid water planning efforts, EPD uses "snap-shot" views of current active irrigation. Several of these have been completed and reported for various planning purposes. Each usually covers only one portion of the state, and each represents a point in time, just as this Ag Water Demand Baseline represents the entire state for fall, 2007 to spring 2008.

EPD-led or cooperative efforts began in the late 1990's as scientists with the J.W. Jones Ecological Research Center, University of Georgia's NESPAL, and EPD's employees sought data to assist planners and negotiators working on the ACF Compact negotiations. The efforts were added to as the Flint River Drought Protection Act (2000) triggered a need for irrigated area on which to make payments that began with the drought in 2001. Additional areas were mapped during scientific studies underlying the preparation of the Flint River Basin Water Development and Conservation Plan (2006) and the Coastal Georgia Water and Wastewater Permitting Plan for Managing Salt Water Intrusion (2007). Most of the mapping labeled here as EPD resulted from farmer involvement. Farmers holding agricultural permits identified their irrigated fields on computer screens, and UGA and EPD mapping specialists drew in and labeled those fields in GIS. County Agents of the Cooperative Extension Service assisted in scheduling and coordinating these farmer visits. Irrigated fields mapped by EPD were tagged with permit source information. Farmers updated EPD data on location of those wells and pumps.

EPD began coordinating its mapping efforts with those of Soil and Water Conservation Commission in the fall of 2006. By then EPD had mapped 21,257 fields in Georgia. For all of these fields, the water sources were also identified. Coverage tended to be most complete in the Flint River Basin and in the Coastal Zone counties since it was those problem areas that brought about efforts to accurately define irrigated area. Other areas were mapped in response to and in cooperation with Cooperative Extension Service "County Mapping Days." This was an opportunity for farmers to visit with EPD and UGA personnel concerning their permits and applications for permits, as well as to identify locations of existing water sources and irrigated fields. Several counties in the center of the Coastal Plain (Wilcox, Ben Hill, Dodge, Telfair, Brooks, Cook, and Berrien) and most counties in the Piedmont, were not involved in these EPD mapping efforts. Taken together, these EPD irrigated field maps covered 1,124,000 acres.
Locations of EPD ag water withdrawal permits and applications for groundwater (blue) and surface water (green) - fall 2008. Statewide coverage EPD-mapped irrigated areas as of fall 2006 when EPD ceased most field mapping.

Close-up of an area in Colquitt County showing locations of several surface (green) and groundwater (blue) permitted pumps and wells.


Close-up of mapped field areas in Mitchell County showing locations of several surface (green) and groundwater (blue) permitted pumps and wells and wetted areas of the fields that they irrigated. Note these are not the permitted maximum area, but the field areas actually in use at the time of mapping.


Close-up of mapped field areas. In these images circular and pie-shapped areas are fields irrigated with center pivots. As is evident in these images, in Georgia, many center pivots cannot be operated in full circles because of roads, trees, utility poles, and property lines. In others, efficient use of the field requires overlapping circles.

SWCC-led irrigation mapping - Meter Locations and Associated Irrigated Areas:



Sample SWCC and ASU mapped fields are outlined in red in the images above and below.
As the Georgia Soil and Water Conservation Commission (SWCC) prepared for and installed water meters on agricultural fields, they contracted for in-field mapping of those irrigation systems and associated water sources. In most cases, the contractors, employees of Albany State University's (ASU) Flint River Planning and Policy Center (FRWPPC), used GPS units in the field and transferred position data from them to create layers in GIS. Those drawings outlined field area and labeled them with type irrigation hardware and source of permitted withdrawals.
Statewide coverage SWCC-mapped irrigated areas as of fall 2008.


In initial mapping contracts ASU was given a list of locations where meters had been installed. They visited each site, mapped the meter location, well or pump locations, and fields that were associated with those installed meters. In subsequent contracts, ASU and their subcontractors at UGA were given a list of permits to track down. With the help of Cooperative Extension Service and USDA's Farm Service Agency and Natural Resources Conservation Service, locations for most of these were found. For permits deemed to be active, field areas that were being irrigated were mapped. SWCC began installing meters for these systems when they received the maps. For inactive permits only the source location was mapped.
At the start of the Ag Water Demand project, SWCC provided maps of all areas then completed. We verified these comparing them with maps originally prepared by ASU-FRWPPC, and against 2007 field images from the National Agriculture Imagery Program (NAIP) to correct for fields that were reconfigured by farmers since original mapping in 2004 through 2006. SWCC mapping included 13,700 fields covering 790,000 acres. There were 10,451 fields irrigated by center pivots; they covered almost 660,000 acres.

Reconciling SWCC and EPD maps:

When EPD and SWCC maps are overlaid, there was considerable overlap. Ideally, these should overlap 100%. But differences in mapping priorities, changes in irrigation systems between EPD and SWCC mapping times, and use of field visits by surveyors using GPS equipment, as compared to office visits by farmers, resulted in differing coverages and different numbers of mapped fields.

Beginning in fall of 2006 EPD and SWCC began working together to reconcile these different coverages. For permits where each had mapped the same irrigation system supplied by the same sources, the field-verified SWCC map was retained and EPD map removed. We continued this same process using geographic tools in ArcGIS. First, we adjusted older EPD field areas that were mapped in a slightly different coordinate system so they better aligned with the modern aerial imagery. Whenever any centroid (center of a circle commonly) of an EPD mapped field fell within a SWCC-mapped field, we removed that EPD shape and retained the newer SWCC field shape.

All SWCC mapped shapes were retained, but there were still many fields mapped with farmers help left in the EPD set of maps. These included maps for fields that SWCC judged as inactive, and fields that for one reason or another were overlooked. Often the latter was because SWCC and EPD lacked current farmer contact information. The rate at which meters had to be installed and the rate at which new irrigated areas had to be mapped caused some omissions, as well.

The enabling legislation for irrigation water metering ordered data to be summarized along hydrologic boundaries - sub-basins (HUC 8 boundaries), and installation proceeded sub-basin by sub-basin. EPD permits and most farm records were kept in data with county boundaries. This presented a challenge in finding all permits that fell in a sub-basin. Finally, flow meters were only installed on permitted systems. EPD maps included many applications for new permits. SWCC would only visit these after farmers had installed their own meter, and mapping would have to wait until then. At this point EPD maps remaining in their layer still included 11,272 fields covering 422,400 acres.
Overlay of SWCC and EPD coverage, Colquitt Co.

Close-ups illustrate overlap, misalignment, and missing maps that required reconciling.

Reconciled SWCC and EPD for an area shown above in SWCC maps).

Previously unmapped center pivots (purple areas in these images) were visible after SWCC and EPD maps were merged and reconciled.

Adding unmapped and new center pivots:

As we reconciled the SWCC and EPD coverages, we quickly realized that there were also unmapped irrigated areas, as evidenced by center pivots that were visible in the 2007 NAIP aerial imagery. To establish an approximate base area for irrigation and for identifying soils commonly found on irrigated fields in each county, we scanned all unmapped fields in the state. The resulting maps allowed others of the Ag Water Demand team to proceed as we worked on closer examination of all mapped and unmapped areas. In this scan we split the state into quarter quads and examined the 2007 imagery for signs of center pivots. These signs included circular fields, pivot tower wheel tracks, pivot pipes, and areas wetted and greened in circular patterns. Because of the high resolution of the 2007 NAIP images (1 meter positional accuracy, but most shapes over 6 inches were easily seen) we could zoom in on suspected pivot fields and verify the pipe and towers. Through this rapid scan, we found 2,000 center pivots that were not in either EPD or SWCC maps. These were labeled as 'UGA' to distinguish their data source.

Field-by-field corrections:

With a preliminary mapping of irrigation covering all counties in Georgia complete, we worked for three months carefully examining every mapped shape and the fields that surrounded them. Each EPD, SWCC or UGA mapped area was compared with NAIP imagery of 2005, 2006, and 2007 and, where available, higher resolution Google Earth imagery. When the field views of a mapped area showed that the area was no longer used in irrigation, the field shape was removed or modified. Evidence included plantings of pine trees, construction of homes and farm structures, un-plowed and unplanted fields, conversion to hunting alleys, or three years of unimproved pasture. Most of the fields that were removed were in older maps of EPD. Often these included corners of center pivot fields that farmers once attempted to irrigate, but have now eliminated them. Many fields that had been identified previously by farmers as irrigated fields now show little evidence of current irrigation. However, unless these areas were small, odd-shaped, or isolated from other irrigated field areas, these cropped or plowed fields were retained. Even if they are not routinely irrigated, they will become the area first irrigated again as prices change, land is rented to other farmers, or new systems are purchased.

Previously mapped center pivots, now in planted pine trees, were removed from EPD map records.
Misaligned and incorrectly drawn EPD pivots (yellow) were redrawn (green).
Many field shapes were redrawn. This was particularly true of older fields in the EPD record, but it even included SWCC mapped fields. In many cases, farmers have reconfigured the irrigation field or installed new hardware since the area was originally mapped. They may have cleared part of a field area to allow a part circle pivot to cover more land. In others they replaced traveler-irrigated areas with center pivots. In some areas that used towable pivots to cover two or more fields, permanent pivots had been added.

In other cases, field areas had simply been drawn inaccurately. Some field drawings were done by hand in a preliminary drawing, but left uncorrected. Others imported or entered incorrect location data or pivot length data.

Whenever an EPD- or SWCC-mapped field was changed, a note was added in the attribute record associated with the drawing in GIS. It recorded the type of change made -- enlarged area, decreased area, realigned field edges, replacement of a traveler for a pivot, etc. When a hardware change was made, the new hardware was noted in a record for hardware type. If earlier maps had left off the hardware type, it was added.

After cleaning up this data source, retained EPD maps provided locations of 8,900 irrigated fields that were not present in the SWCC metered field maps. These covered 370,000 acres of row crops and orchards. These EPD areas also included 2,410 center pivots that irrigated 152,000 acres.
This SWCC mapped towable pivot (red) was realigned to correct borders( green).

Since original mapping, this SWCC mapped traveler field (red) had been converted to a center pivot (green).

These EPD-mapped pivot and traveler fields (yellow) were converted for drip irrigation (green).
Originally drawn as a traveler field, this EPD mapped traveler field (yellow) had been converted and was redrawn as a center pivot (green).

Traveler field in 2007 NAIP imagery
Center pivot seen in GoogleEarth imagery
Center pivot in 2007 NAIP imagery
Drip fields in GoogleEarth
Four new pivots

New field areas were added during this field-by-field examination as well. Closer inspection often revealed center pivots that earlier scans had missed. Likewise we found evidence for traveler-irrigated fields, and they were drawn in. That evidence included rows of risers placed 150 to 250 ft apart across a rectangular or irregular field. Risers bring the water into the field from nearby water sources through buried pipe. Traveler hoses are attached to them and then traveler gun cart and hoses are pulled from then to start each pass of the irrigation. Traveler fields could also be noted by cleared or pushed-down plants in alleys where the hoses were dragged. In some fields patterns of wetted soil or greened field areas among drought stressed field areas highlighted irrigated areas. The 2007 imagery, taken during a widespread drought in Georgia, made these areas very evident. In other cases traveler hardware, hoses, or sprays from big guns could be seen in the fields.

Drip irrigated vegetables were seen in many of the aerial images. Distinctive patterns of fields including manifolds delivering water to small irrigation zones and crop rows, alleys left for spray or harvest equipment, and plastic mulch covering parts of beds were common clues. Unfortunately many of the aerial images were taken in late summer or early fall. It is likely that we missed drip fields that had been irrigated in spring but were already plowed. However, we attempted to use multiple image sources to aid in their identification. In a crop type record in the attributes table, we recorded these as 'vegetable' areas.

Orchards and some perennials, including pecan, peach, apple, grape, and blueberry were usually readily identified in aerial images. What was not so evident was whether or how they were irrigated. We used a few guidelines to distinguish between irrigated and non-irrigated perennial crops. The primary one was the presence of a permitted or proposed (permit application) water source within or adjacent to the field. Most peach and blueberry areas and many pecan areas had these wells and surface pumps mapped, but no field was drawn for the application area. New trees, replanted trees, and uniform growth in orchards were evidence of an irrigated orchard. Conversely, old orchards with many missing trees, houses within small orchards, etc., were taken as evidence of no irrigation. Peach orchards and blueberry fields were plainly visible and were mapped, but irrigation was assumed to exist on only 25% of irrigated peach orchards (Kerry Harrison, personal communication). Usually those orchards had permitted withdrawals within or adjacent to the orchards. When hardware had not been noted by EPD or SWCC for orchards, it was listed as 'SS, DR' indicating that it was either solid set sprinkler or drip/micro spray, the two prevalent irrigation methods in orchards and groves.

Sod production and container and in-ground plant nurseries were also individually mapped. Often these have been noted in SWCC and EPD maps as well. Harvest patterns, often seen under center pivots, identified many of these sod farm areas, although undoubtedly some sod farm areas were overlooked as they were seen in winter images that were less distinctive. In a crop type record in the attributes, we recorded these as 'turf' or 'nursery' as appropriate.

Golf courses and atheletic fields were permitted as agricultural areas in some cases. In all cases we eliminated these from agricultural production areas. Irrigated area of golf courses was not included in the Ag Water Demand study.

In all, 4,676 addtional irrigated fields were mapped through visual inspection. These included 2,973 center pivots that were visible in 2007 imagery but not included in SWCC or EPD maps. Of the remaining, 757 were traveler irrigated fields and 620 drip irrigated fields. The effort added 210,000 acres, including 136,000 acres in center pivot land, to the State's inventory of irrigated farm land.
Statewide view of UGA-added irrigated areas, March, 2009.

Clinch County blueberry fields.
Established pecan orchard.

Projecting crop type onto mapped fields for Baseline Irrigated Areas:

For some mapped field areas -- perennials or special-use fields -- crops could be identified specifically from aerial images or field notes of mapping crews. These included permanent orchard and nursery crops, sod farms, drip irrigated vegetables (as a group), and multiyear irrigated pastures. However, specific field crop designations could not be made for most crops simply on the basis of 2007 imagery. We labeled these 'rotation' fields, whether they were in peanut, cotton, sprinkler irrigated vegetables, winter cover, or even hay at the time of 2007 images. Still, as a starting point for UGA's projections of irrigated crops in 2011, 2020, 2030, 240 and 2050, we needed a baseline of irrigated cotton, peanut, corn, etc. To provide that baseline, all fields that were involved in rotations were added together for a county. Then 2008 Cooperative Extension Service Irrigation Survey results were used to apportion acres among mapped areas of these rotation fields.

Most agronomic row crops are produced rotationally - crops change from year to year on the field. While a pivot may remain in place year after year, the crops grown under it may change each season, often more than once per year. Peanut, cotton, corn, and soybean make up most of these fields each year. Wheat, rye and other winter grain or cover crops may appear on them at certain times of the year, but those receive little water in most years. Vegetables like sweet corn, snap beans, onions, carrots, watermelon and canteloupe are also produced in rotations with these agronomic crops.

The 2008 CES Irrigation Survey was the latest of surveys of irrigation taken every three to four years. Based on their experience in their county, agents report which crops were irrigated and they estimate acres of each and water amounts added for the survey year. We added the acres reported for corn, cotton, peanut, soybean, tobacco and sprinkler irrigated vegetables to provide a total of survey rotation acres. The ratio of a crop, e.g. corn survey acres to rotation survey acres was then multiplied by that county's mapped rotation acres to get the baseline crop acres, in this case for corn. Baseline acres for these rotation crops plus specifically mapped irrigated pecan orchards were then used in the economics models when projecting changes in acres for irrigated crops through 2050.


Last updated 7/20/2009 James E. Hook