TERM PAPER

CEE 6440 GIS IN WATER RESOURCES

FALL 2007

BUILDING AN ArcGIS DATABASE FOR MILLARD COUNTY, UTAH AND AREA SUITABILITY ANALYSIS

INTRODUCTION

Agriculture is an integral part of any economy, since our food needs are tied to it. Managing water use in irrigation is a potential concern. High energy costs and water shortages need to be recovered, by acquiring knowledge of yield losses due to limited water resources. Currently most of the farmers irrigate the fields based on experience, visual inspection or experimental results, or on the basis of water availability, regardless of whether the crop needs it or not. Such methods are very site-specific and cannot be used to make decisions or long-term forecasts.

To begin with in this project we focus on building a database consisting of layers such as DOQ’s, soils, hydrologic structures in the area, ETM (thematic mapper) land cover images and some information of the agriculture grown in the area.

PROBLEM STATEMENT

The area selected for the project is Lower Sevier River, Millard County, Utah. As can be seen from the following figure, most of the land in Millard is occupied by agriculture (Land use in 2003 from AGRC).

Most of the agriculture in the Millard County is concentrated in Delta as can be seen from the following figure.

The land cover is distributed between Alfalfa, grain crops and riparian vegetation, rest is either fallow or occupied by water sources like reservoirs or canals.

If we look at the important vegetation, alfalfa comprises almost half of the agriculture in the area.

This explains the importance of the area and why many agricultural experiments are being conducted in the region. Hence the objectives of the study are outlined below.

OBJECTIVES

· To study the typical crop growth cycle of the main crops in the area chosen to optimize the timing of the images.

· To select an irrigated area in the Sevier River basin.

· To build a GIS database in ArcGIS containing layers such as DOQ’s, soils, hydrologic features, field boundaries, crop distribution and land use, and TM satellite imagery.

· To examine the areas that have good yearly crop growth and poor yearly crop growth, and compare to the soil types etc.

LITERATURE REVIEW

According to the literature review, Utah Spring wheat is planted in between March 20 and May 1 and active planting period is from April 1 to April 20. The harvest dates are usually from July 10 to Sept 10 and the most active harvest period is from Aug 05 to Aug 25. Whereas Utah winter is harvested from July 10 to Sept 01, the most active period being from July 25 to Aug 10. For alfalfa, the surveys show that in Utah, the harvest or cutting of Alfalfa crop is usually between Jun1 and Oct 25.

So roughly, the images for wheat should be roughly between March to Sept. For alfalfa we should be looking at images before June specifically (if we need the total biomass of the crop).

The site selected was Delta, Millard, Utah.

DATA SOURCES

There are a lot of websites which offer free GIS data but the most valuable ones for this study were:

1. Automated Geographic Reference Center (AGRC) – The layers of field boundaries, DOQ’s, streams, hydrologic features, agricultural land use in 2003, wetlands, lakes and watersheds (refer to Appendix) were collected from this source.
2. Intermountain Region Digital Image Archive Center (IRDIAC) – The LANDSAT ETM+ (Enhanced Thematic Mapper) images were collected from this site. There were only 3 images which corresponded to the irrigated area, Delta. The dates on which the images were taken were April, 26 and July, 31, 2000, and Nov, 02, 1999.
3. Natural Resources Conservation Service (NRCS) – The soils data was collected from NRCS. The data comprised of the soils classification based on various soil properties discussed in the results.

METHODOLOGY

The following steps were performed for the project:

1. The layers for DOQ’s, soils, field boundaries, land use etc were downloaded in .shp format. The projection for the layers was NAD 1983 UTM Zone 12N.

2. The ETM+ images from the IRDIAC site were in Albers Equal Area Conics projection and in .tif format. Hence to overlay and study the layers, projection of TM images was changed to NAD 1983 UTM Zone 12N. This was done as follows:

Select ArcToolbox – Data Management tools – Projections and Transformations – Raster – Project Raster – Select Input Raster (to the .tif image) + Output Raster (to a specific file) + Output Coordinate System (Projected Coordinates – Utm – NAD 1983 – NAD 1983 UTM Zone 12N.prj)

3. Once the projection is changed the image can be overlaid with the soils layer for analysis.

4. A similar change of projection was done in ERDAS Imagine. The results from this analysis are also presented.

ANALYSIS

1. Using ERDAS Imagine

ERDAS Imagine was used to conduct some analysis to identify vegetated areas. The pixel values are the best indicator for the amount of vegetation since vegetation shows big value in the NIR band or the red regions in the images. The resolution of the image was 30 m, hence it could detect most of the land features quite well. High pixel values in red (NIR) showed dense vegetation, low pixel values indicated riparian zone or water. Large irrigated areas were identified since they were irrigated using central pivoted systems which could be differentiated in the image quite easily. ERDAS image analysis showed that lot of the vegetation growing in the region was very dense.

2. Using ArcGIS

The hydrologic features layer showed a lot of canals in the region which confirms that the area is irrigated on-demand.

The soils layer was overlaid on the images and the soil types were studied from the region. The following soil types were identified which had vegetation, from the tables available with the images.

Aa	482372	Abbott silty clay	Poorly drained
Ab	482373	Abbott silty clay, strongly saline	Poorly drained
Af	482374	Abbott silty clay, sandy substratum	Poorly drained
Ag	482375	Abbott silty clay, sandy substratum, strongly saline	Poorly drained
Ah	482376	Abraham loam	Somewhat poorly drained
Ak	482377	Abraham loam, strongly saline	Somewhat poorly drained
Am	482378	Abraham silty clay loam	Somewhat poorly drained
An	482379	Abraham silty clay loam, strongly saline	Somewhat poorly drained
Ar	482380	Alluvial land, wet	Poorly drained
As	482381	Anco silty clay loam	Somewhat poorly drained
At	482382	Anco silty clay loam, strongly saline	Somewhat poorly drained
Av	482383	Anco silty clay loam, sandy substratum	Somewhat poorly drained
Pe	482408	Penoyer silt loam	Well drained
Ph	482409	Penoyer silt loam, strongly saline	Well drained
Po	482410	Poganeab silty clay loam	Poorly drained
Pr	482411	Poganeab silty clay loam, strongly saline	Poorly drained
Pt	482412	Poganeab silty clay loam, sandy substratum	Poorly drained
PA	482413	Pahranagat loam	Very poorly drained
PM	482414	Playas	Very poorly drained
PN	482415	Playas-Abbott association	Very poorly drained
PU	482416	Poganeab-Uffens-Uvada association	Well drained

Looking carefully at most of the images, the dense vegetation in the region corresponded to silty clay loam soil type primarily. The soil salinity was identified from other layers.

The vegetation patterns in the images were studied on the same patch of the area and some interesting features were noticed. The first set shows that “Am, An” soil types have vegetation till July whereas vegetation in the Penoyer silt loam (Pe) soil diminished in July image and same can be seen in the last image of Nov.

To validate this, a second set of vegetated area was studied. “Am, An” soil types supported vegetation from April to July but on the other soil types the vegetation became sparse.

Looking at the poorly vegetated areas, “PM, PN” were the dominant soil types in the images where there was hardly any vegetation.

RESULTS

The analysis and some information provided in form of .txt files were utilized to come up with some results. Some more analysis was done to find out the contributing factors for vegetated soils. The numbers shown below were categorized into various criterion (as shown below) but separate tables were not provided. It was found that the soil types which had vegetation corresponded to the ones which had high alluvium and had a lot of humus content in the surface layer which makes it obvious why they would support vegetation vs. others which were characterized by lacustrine deposits.

Alluvium	Aa	482372
Alluvium	Ab	482373
Alluvium	Af	482374
Alluvium	Ag	482375
Alluvium	Ah	482376
Alluvium	Ak	482377
Alluvium	Am	482378
Alluvium	An	482379
Alluvium	Ar	482380
Alluvium	As	482381
Alluvium	At	482382
Alluvium	Av	482383

These soils also had low cobble content (15%), which means it is easy to till the top layer for agricultural purpose. Though the soil types suggested that the soils supporting vegetation had “strong salinity” but it was found that the soils not having vegetation were actually saline. Infact the electrical conductivity (EC) of the surface layers of the soils having vegetation was 4 to 8 mmhos/cm which was reasonable but could not be categorized as highly saline which had a limit of EC of 8 mmhos/cm. Since, the soils with high salinity have reduced available water capacity and may induce salinity toxicity, which restricts vigorous plant growth and re-establishing vegetation in disturbed area. The soils also had high adsorption which means a high Carbon exchange capacity (CEC) to clay ratio to a depth of 50 cm (root zone) which supported vegetation. The other interesting observation was that high sodium adsorption ratio has the potential to restrict plant growth and the content in the soils was found to be somewhat limiting (> 3 to =< 13).

SUMMARY & CONCLUSIONS

The study came out with lots of interesting things. Delta, Millard was found to have dense irrigated vegetation. The central pivoted irrigation system was used in some parts of Millard. The region was dominated by Alfalfa and grain production. About the soils supporting vegetation, it can be concluded that poorly drained wet silty clay loam soils with moderate salinity were the best for agriculture in the region. This seems a bit strange because salt can inhibit plant growth by creating a crust on the soil, but might be preserving the moisture within the soil and hence supporting vegetation, since Delta is an arid region with most of water coming from irrigation. Also the soils were wet or poorly drained ones but can sustain vegetation since they can hold water for long time specially for a grass species like Alfalfa. Also the soils type in the regions matches with the USDA NRCS soil surveys, which show that for Alfalfa, sandy loam to clay-loam soils are preferred.

RECOMMENDATIONS

1. It is hard to get data from the MODIS website, some tool would be useful if it has most of Utah images. It would be good to validate the results using MODIS images but they are hard to find for some remote regions.

2. TM images are very limited for the region. It would be good to analyse more images to come to a conclusion on what soil types are affecting agriculture in the region.

3. It was hard to make a mosaic of the DOQ’s for the county since the image was losing its resolution. This could be improved by making a DOQ from a DEM.

REFERENCES

http://gis.utah.gov/component/option,com_dbquery/Itemid,87/

http://earth.gis.usu.edu/

http://soildatamart.nrcs.usda.gov/