Due Week 13

                (photos - pfh: soil erosion and gullying in the upper Panuco basin, Sierra Madre Oriental, eastern Mexico)

PURPOSE

The purpose of this lab is to apply raster GIS methods to a soil erosion model for a drainage basin displaying varying types of topographic and land use conditions.  Two sections of a large watershed in the headwaters of the Panuco basin are utilized. Specifically you will be examining the Rio Tempoal basin, located in the eastern flanks of the Sierra Madre Oriental in Veracruz, Mexico. You will combine a variety of skills learned in earlier labs to a unique method for modeling / simulating soil erosion using the ever popular Revised Universal Soil Loss Equation (RUSLE).

OBJECTIVES

I.  GIS Concepts

• Modelling physical processes with GIS
• Raster (grid) analysis; local operations and neighborhood operations
• Map algebra
• Boolean operators

II.  GIS Techniques

• Raster techniques for creating data subsets from large data sets
• Use of the ArcHydro extension to model watersheds and flow accumulation
• Creation of graphs to summarize spatial relationships
• Spatial Analyst and Raster calculator
• On screen digitizing
• USGS seamless data (SRTM) for Mexico

III.  Geographical

• Watershed definition and delineation
• Contributing area and flow accumulation modelling
• Basin morphometry
• Watershed hydrology
• Erosion modeling
• Land use planning

BACKGROUND

Soil erosion is a concern for farmers, development agencies, and governments throughout the world.  Since the early 20^th century, soil erosion, by wind and water, has been recognized as a major factor for decreases in both soil fertility and land value.  However, soil erosion is a natural geomorphic process.  The mechanisms involved in soil erosion by water vary over time and space. Some of these mechanisms are raindrop splash, unconcentrated down slope wash (sheet erosion), concentrated down slope wash (rill and gully erosion), and a mixed process in which entrainment is by raindrop splash and down slope transport is by surface wash (Luk, 1979).  How significant these mechanisms affect rates of erosion depends on several factors including percent ground cover, soil texture, soil structure, soil porosity/permeability, and topography (i.e. slope gradient) (Rainey, 1991).  Humans can influence the dynamics of each of these and thus, improper human land management can accelerate rates of erosion.

On cultivated land, loss of topsoil due to erosion is a problem (Pimetel, 1983).  In developing nations, the expansion of farming practices and pastures onto remote, steep, hill slopes increases the potential for accelerated rates of erosion (Millward and Mersey, 1999).  Excessive erosion is associated with diverse negative on and off site impacts, including loss of soil nutrients leading to a reduction in crop yields, decreasing stream competence and capacity because of sedimentation, and siltation of reservoirs (Stoddart, 1969).  Additionally, agricultural chemicals transported with soil particles have significant impacts on water quality.

Two types of erosion occur on agricultural land: sheet erosion and rill erosion.  Sheet erosion washes down slope in a uniform manner over a surface area.  Rill erosion is concentrated wash and can lead to gully formation (note above photo).  In general, three techniques are used to measure soil erosion by water: plots and traps, surveys of varying methods, and different types of tracers (Loughran, 1989).  The information on the temporal and spatial distribution of long-term soil loss in drainage basins and on the rates of soil erosion generated by these techniques is used to calibrate and test various soil erosion models (Loughran, 1989).

Soil erosion models simulate the effects of land management activities on rates of soil erosion and contribute to the development of appropriate intervention strategies.  One way to reduce the negative consequences of accelerated soil erosion is to select land use and farm management practices that generate a minimum amount of soil erosion.  Modeling soil erosion provides a sophisticated tool for selection of appropriate soil conservation practices.  There are many soil erosion models, including the European Soil Erosion Model (EUROSEM), the Water Erosion Prediction Project (WEPP), the Limberg Soil Erosion Model (LISEM), and the Chemical Runoff and Erosion from Agricultural Management System (CREAMS) to name but a few.  The most extensively used model is the Revised Universal Soil Loss Equation (RUSLE).  The RUSLE model has advantages because its data requirements are not too complex or unattainable, it is relatively easy to understand, and it is compatible with GIS (Millward and Mersey, 1999).  When used in conjunction with raster-based GIS, the RUSLE model can isolate locations of erosion on a cell by cell basis, determine the role of individual variables on the rate of erosion, and identify the spatial patterns of soil loss within a watershed (Millward and Mersey, 1999). 

Many regions within humid mountainous environments undergo high rates of soil erosion, and this includes parts of Mexico. Other pressures in Mexico that accelerate land degradation and soil erosion are due to increasing anthropogenic pressure on natural resources (Landa, et al., 1997).  Remote, steep, and fragile regions of Mexico are increasingly utilized for agriculture and cattle grazing.  Modeling the sources and outcomes of environmental degradation in these remote regions can help stem the long-term, negative impacts that result from erosion.  In this lab, you will model/simulate soil erosion for parts of the upper Rio Tempoal drainage basin using the RUSLE model and Arc GIS. The Rio Tempoal drains the humid eastern flanks of the Sierra Madre Oriental, and is primarily located within the states of Hidalgo, San Luis Potosi, and Veracruz. The Tempoal is a large tributary to the Moctezuma drainage system, which eventually forms the Rio Panuco before draining into the Gulf of Mexico at Tampico, Tamualipas.

The Revised Universal Soil Loss Equation (RUSLE)

The RUSLE is a revision of the Universal Soil Loss Equation (USLE), which was originally developed to predict erosion on croplands in the United States.  With the revision, the equation can be employed in a variety of environments including rangeland, mine sites, construction sites, etc.  The RUSLE is an empirical equation that predicts annual erosion (tons/acre/yr) resulting from sheet and rill erosion in croplands.  The RUSLE is factor-based, which means that a series of factors, each quantifying one or more processes and their interactions, are combined to yield an overall estimate of soil loss.  It is the official tool used for conservation planning in the US and many other countries have also adapted the equation.  Although scientist at the USDA plan to replace the model with the new Water Erosion Prediction Project (WEPP), the RUSLE is still very relevant and is particularly useful as a teaching tool.  The equation is:

A = R * K * L * S* C* P

where,

A = Annual soil loss (tons/acre) resulting from sheet and rill erosion. This is the predicted value resulting from the execution of the equation above.

R= Rainfall - runoff erosivity factor. This factor measures the effect of rainfall on erosion.  The R factor is a summation of the various properties of rainfall including intensity, duration, size etc.  It is computed using the rainfall energy and the maximum 30 minutes intensity (EI30).  Rainfall erosivity has been mapped for the entire US and most commonly one would simply locate their drainage basin on a map and identify the R value.  Obtaining the R-value for Mexico is a little more difficult.  In this case, we will use a value estimated from a study conducted in Jalisco, Mexico (see Millward and Mersey, 1999).

K= Soil erodibility factor.  The soil erodibility factor measures the resistance of the soil to detachment and transportation by raindrop impact and surface runoff.  Soil erodibility is a function of the inherent soil properties, including organic matter content, particle size, permeability, etc.  Because these properties vary within a given soil, erodibility (K values) also varies.  Erodibility values of the major soils in the US can be obtained from the County Soil Surveys Reports.  For Mexico, one study generated these values using information from the Secretaria de Agricultural y Recursos Hidraulicos (1991) and an INEGI soil map of the area (www.inegi.gob.mx).

L= Slope length factor. This factor accounts for the effects of slope length on the rate of erosion.

S = Slope steepness factor. This factor accounts for the effects of slope angle on erosion rates. All things being equal, higher slope values have greater erosion rates.

C = Cover management factor. Accounts for the influence of soil and cover management, such as tillage practices, cropping types, crop rotation, fallow, etc..., on soil erosion rates

P = Erosion control factor. Accounts for the influence of support practices such as contouring, strip cropping, terracing, etc...

Once these factors have been determined for a field of interest A can be computed. Also, the equation can be used to determine the desired cover management factor (C ) or erosion control (P) if the allowable soil erosion rates are known. Thus, you can use the RUSLE to simulate the impact of changes in land use and land cover on soil erosion... anthropogenic impacts on the environment!!!!

REFERENCES

Avwunudiagba, A. and Hudson, P.F.  A review of soil erosion models with special reference to the needs of tropical mountainous environments.  Unpublished manuscript.

Landa, R., Meave, J., and Carabias, J. (1997). Environmental deterioration in rural Mexico: an examination of the concept.  Ecological Applications 7(1), 316-329.

Loughran, R.J. (1989). The measurement of soil erosion.  Progress in Physical Geography  13, 216-233.

Luk, S.H. (1979) Effect of soil properties on erosion by wash and splash.  Earth Surface Processes 4, 241-255.

Millward, A.A. and Mersey, J.E. (1999). Adapting the RUSLE to model soil erosion potential in a mountainous tropical watershed.  Catena 38, 109-129.

Pimetel, D. (1993).  Overview.  In: World Erosion and Conservation, Pimetel, D. (ed.).  Cambridge: Cambridge University Press.

Rainey, S.J. (1991). Cultivators in the clouds: a study of erosion in relation to agricultural practices in the Sierra Alta of Hidalgo, Mexico.  Masters Thesis, The University of Texas at Austin.

Secretaria de Agricultura y Recursos Hidraulicos.  (1991).  Manual de Prediccion de Peridas de Suelo por Erosion.  Colegio de Postgraduados, Guadalajara.

Stoddart, D.R. (1969). World erosion and sedimentation.  In Water, Earth, and Man, Chorley, R. (ed.).  London: Methuen, 43-64.

Internet Resources to Soil Erosion

• USDA RUSLE
• Soil Erosion Links
• Other Soil Erosion Links

Part I:

Generating the RUSLE factors for areas within the Rio Tempoal Drainage Basin

In this section you will build a GIS data base of the RUSLE factors (R, K, L-S, C, P) for an area of the Rio Tempoal Drainage Basin using ArcGIS.  This will enable you to simulate erosion in the drainage basin.

• in this lab you will make use of several techniques utilized in previous labs which are not reviewed in Lab 9.

1.  Download the DEM of the study area in Mexico.  To help locate the area, use the topographic map hand-out.

• Open USGS Seamless Data Distribution System

• Enter as an International Viewer

• Zoom to and Download a SRTM 90m image of the area (Fig. 1) (Note: only 90m pixels are available for international data in the USGS seamless archive...and this has only become available recently.) 

• Save the .zip file to your class storage folder and open it.

• Define Projection for the DEM: 

  • The DEM projection is WGS84, later in the lab (specifically, when you find the slope of the Drainage Basin) you will have to use a projection that is compatible with elevation measured in meters.  Therefore, reproject the grid to UTM.

    •   Open Arc Toolbox, go to Project Wizard (coverages, grids).  Choose the DEM you just opened in WinZip as the Input Dataset.  Select UTM, then: units = meters, and zone = 14, and set the Datum to NAD 1983 - Mexico and Central America, save the output in your storage space as Area_UTM.  Under Grid settings, check the "Project will calculate cellsize..." box and specify 90 meters.

2.   Copy the Data files into your directory.

• Use Arc Catalog to copy the following shapefiles to your class storage space:  R_Factor.shp, Tempoal_soils.shp, C_Factor.shp, and P_factor.shp.

• Launch Arc Map.

3.    Delineate the Drainage Basin.

• Use Editor (use the names: Basin_bnd, basin_limit), Spatial Analyst (use the name: tempoal_basin), and Arc Hydro (use the names: tempoal_catchments and tempoal_streams) to delineate the drainage basin (Fig. 2) and (Fig. 3). (Remember how to do this from way back in Lab6?)

4.    R-factor:

• A problem with the R-factor in our analysis is the pronounced wet-dry precipitation regime of the Upper Rio Tempoal.  The average precipitation from May to October ranges between 2000 mm and 2300 mm and 400mm to 500mm from November to April.  Thus, the total annual precipitation is 2600 mm ( 2150mm + 450 mm).  As is common to many areas within the developing world, the rain gauge data (mm/day) in this area is not suitable for deriving precipitation intensity (mm/hr).  Therefore, you will analyze the erosivity factor on an annual basis.  Using a second order regression equation you can estimate erosivity values using the value for annual precipitation.  The R value is in MJ*mm/ha*h*yr.

                    R = - 0.0334 (P annual) + 0.006661 (P annual) Squared

• Load the R-factor.shp. Open the attribute table of the R-factor.shp.  Add a new field labeled R_factor.  Calculate the R-Value for the drainage basin. Enter this value in the new column of the R-factor table.  Now convert the R-factor Shapefile to a Grid by highlighting on the R-factor shapefile in the table of contents and going to Spatial Analyst > Convert > feature to grid.  Select R_factor for the field.  Name your grid R_factor.

5.    K-Factor:

• The soil erodibility factor (K) represents the average long-term soil and soil-profile response to the erosive power associated with rainfall and runoff.  An INEGI map of the area provides a classification scheme.  Each soil type has a distinct K-value associated with it.
• Load the Tempoal_soils.shp and examine the attribute table.  Note the following attributes of the Tempoal_soils: Soil_type = the various soil types in the Drainage Basin, K-factor = the erodibility index of the various soils.  Open the Tempoal_soils.shp Properties and go to Symbology. Choose Categories > Unique Values and K-factor as the Value field and Add all values (Fig 5).  Now convert the Tempoal_soils.shp to a grid. Choose the K_factor as the attribute to use. Follow the procedure you used for the R_factor. Name your new grid K-factor.

6.    LS-factor:

• In executing the RUSLE, the L and S factors are combined into the LS factor. The LS factor can be measured directly in the field or from USGS quadrangle maps.  Here you will generate the LS factor from the DEM. Before you delineated the Drainage Basin, you created a mosaic of the DEM called Tempoal_Basin.  This was a rough estimate of the Drainage Basin.  You will only work with the portion of the DEM within the actual Drainage Basin which you labeled Tempoal_Catchments. Therefore, you will need to make a mask of the area of the Drainage Basin out of the DEM.  To do this, use some of the techniques you used in earlier labs and earlier in this lab (hint: labels you will use are: Basin_Raster and Demsub). Subset the DEM so that you have only the portion within the Drainage Basin. 

• Now you will compute the LS factor using the Demsub and GIS.  With Demsub as your active theme go to Spatial Analyst > Surface Analysis  and use the Slope tool to calculate a slope surface for the area.   Name your result Tempoal_slope . Next you need to derive the flow accumulation one of the inputs required to compute the RUSLE.  Earlier, you created a flow accumulation (Fac) when you delineated the Drainage Basin in Arc Hydro and you can make a mask of this using Basin_Raster as the mask for Fac.  This will give you the flow accumulation for the actual Drainage Basin. Name the new theme flowacc (Fig6).

• You have all the themes necessary to generate the LS factor using the DEM.  Using Raster calculator build the following expression.  (Note: the value for resolution corresponds to the cell size of the DEM.) What is this cell size of the DEM? Use this value for the resolution in the expression below:

1.6 * Pow(([flowacc] * resolution) / 22.1, 0.6) * Pow(Sin([tempoal_slope] * 0.01745) / 0.09, 1.3)

 

• Name the new grid LS.

7.    C-factor:

• Load the C-factor.shp, examine the attribute table, and change the symbology.  Note that the C-factor was derived from landuse/ land cover types.  The land use types are  Bosque Mesofilo de Montana (Dense Pine Forest), Selva Alta Perennifolia (Hardwoood Forest), Pastizal Inducido (Intensive grazing or agriculture), and Selva Alta Perennifolia y Agricultura Temporal (Hardwood forest and seasonal agriculture).  Now, convert the C-factor shape file to a grid. Make sure you select the C-factor as the attribute to use in generating the grid.  Name the new grid C_facgrid.

8.    P-factor:

• Load the P-factor.shp and convert to grid following the procedure you adopted in generating the C_facgrid. Name the output P_facgrid.

9.    Run the RUSLE equation:

• Now you have all the factors necessary for executing the RUSLE. Use the map calculator to simply execute the following expression:

([ R-factor]*[LS]*[K-factor]*[C-facgrid]*[P-facgrid]) (Fig 8)

• Name the result Soiloss. This is the annual soil loss (tons/acre) for the Tempoal drainage basin.

Part II

In this portion of the lab you will simulate (model) soil erosion in the Drainage Basin under different land use and land cover change scenarios. You are part of a research team charged with devising a suitable soil conservation strategy for the watershed. Once again, you will be utilizing skills learned from previous labs.

Section 1: Modeling soil erosion to changes in land use and land cover.

• Scenario 1
• As is common throughout Latin America, small scale farmers who live in the watershed will be forced to emigrate to cities in order to find work, some of these farmers are Indigenous Huasteca.  In this scenerio, all of the area within the Basin will bought by a wealthy rancher from nearby Tamazunchate, who plans on turning the entire basin into a cattle ranch.  After the area is cleared, it will remain bare for a large part of the year, and only after several months will grass begin to grow. Simulate the soil erosion for the Drainage Basin that will occur from this dramatic disturbance. Assume that R, LS and, P factors remain constant. Use a C-factor value of 0.90 for the mostly bare soil in that first year.

 

• Scenario 2

• The cattle ranch in Scenario 1 has failed.  Beef prices declined exponentially due to a new virus.  However, corn and grain have shown a steady incline in prices.  So the land owner has decided to convert the area back to a mixture of fallow and agriculture involving the cultivation of several types of crops. Again, simulate the erosion for the Drainage Basin resulting from the change in surface characteristics. You will utilize the information provided in the table below to generate a new C-factor for your simulation. Again, assume that R, LS and, P factors remain the same.
  • Using the feature ID as a unique #, edit the attribute table for the C-factor shapefile (located on your network directory) and include crop type and appropriate values for the corresponding C-factor

 

FID	Crop type	C-factor
3	Corn	0.850
0	Forest	0.001
2	Sorghum	0.520
1	Alfalfa	0.100
		* estimates

• Scenario 3:  

• Now, none of the above scenarios ever took place.  The local small scale farmers remain in control of the land and have increasingly converted the entire drainage basin to shade coffee cultivation.  It has been argued that this is a very low intensity land use.  Using the below values for the indicated cells, repeat the procedure from scenario 2 for the drainage basin under shade coffee cultivation.

 

FID	Land Use	C-factor
2	Selva Alta	0.03
0	Bosque Meso	0.001
1,3	Shade Coffee	0.003
		* estimates

Section 2: Spatial analysis of varying soil erosion rates

• Use map calculator to create GRIDs illustrating the following
  • total amount of soil erosion for area as 1) a cattle ranch, 2) reversion back to agriculture, 3) shade coffee cultivation
  • the difference between the three scenarios

Section 3: 

• Write a 5 page report detailing your findings and recommendations for land management of the Tempoal Drainage Basin.  Be sure to compare the three scenarios from section 2.  Create Layouts for each of the above scenarios and refer to them in your report.

Grading:

• A map layout (with all cartographic elements) of the Soilloss created in Part I.
• A map layout (with all cartographic elements) for the six GRIDs you created in Part II of the lab (3 from section 1, 3 from section 2)

• A five page (double spaced, 12 pt. font, 1" margins) report that addresses the following:
  • where is the greatest amount of change in erosion rates from the "natural" to the cleared surface?
  • considering the change in erosion rates following removal of the "natural" vegetation, which portions of the Drainage Basin would you prioritize for conservation? Why?
  • How much and where in the basin were the greatest and least amounts of changes in erosion rates following the conversion from a cleared surface to agriculture?
  • given the projected erosion rates for the various land use types, which areas should not be allowed to be farmed/ranched?
  • overall, which portion of the basin is most sensitive to land use / land cover change?
  • what are some of the problems with using modeling to analyze soil erosion?

created by aa and pfh, 11/10/02, modified by mf 4/'04, last modified by pfh 11/7/'05