Modeling erosion and deposition pattern for a small watershed

Helena Mitasova, Bill Brown, Lubos Mitas
University of Illinois at Urbana-Champaign

WORKPLACE DOCUMENT

1. Computation of DEM from IFSARE and field data

Location of the study area size of the study area: 154539 square m

IFSARE data processed by TEC

• original 2.5m resolution IFSAR DEM
  

• smoothed DEM 2m resolution
  
• slope
  
• aspect
  
• profile curvature
  
• tangential(contour) curvature
  
• upslope area
  

IFSARE data as recieved from contractor

• raw 2.5m resolution IFSAR DEM with imagery draped over part of it for a begger watershe, the shift in vector data (road, streams is probably error in our database, which we haven't found yet)
  
• raw 2.5m resolution IFSAR DEM with imagery draped over it, 4x vertical exaggerration
  
• raw 2.5m resolution IFSAR DEM with imagery draped over it, no vertical exaggerration
  
• smoothed DEM, 2m resolution
  
• slope
  
• upslope area
  

Field data (measured by GPS)

• interpolated surface with no effective smoothing
  
• given data points and interpolated surface with no effective smoothing
  
• some points are almost 10m above or below the neighboring points
  
• creating weird isolated peaks and pits
  Are these points errors or do they describe some real features not captured in IFSARE data? If these are errors would it be possible to identify the "bad" points and remove them? Or, randomly remove 50% of points and the clen up the rest?
  

• interpolated surface with significant smoothing, 1m resolution
  
• comparison of 1m contours from IFSARE and smoothed DEM from field data
  
• slope
  
• uplsope area
  

2. Erosion modeling

The following data were used for rainfal, soil and cover properties: R=160, C=0.04, K=0.17 on ridge, K=0.32 in the valley

K-factor

Based on smoothed TEC IFSARE elevation data (2m resolution)

• modified USLE LS (not normalized!)
  
• USPED topographic potential for erosion and deposition - note that there is very low potential for erosion or deposition on hillslopes, so I don't expect a lot of variability in Cs, however there is a striking, severe erosion potential in the center of this watershed
  
• USPED topographic potential for erosion and deposition draped over elevation+sediment flow rate surface
  
• SIMWE topographic potential (will be added later)

Based on smoothed contractors IFSARE elevation data (2m resolution)

will be added

• modified USLE LS (fix-normalize !)
  
• modified USLE
  
• USPED topographic potential for erosion and deposition
  
• USPED net erosion/deposition, p=1.6
  
• USPED net erosion/deposition, p=1.6, values in sampling points displayed as the height of cylindrical point symbols
  
• USPED net erosion/deposition, p=1
  
• USPED net erosion/deposition, p=1, values in sampling points displayed as the height of cylindrical point symbols
  
• SIMWE water depth for 36mm/h rainfall intensity, uniform conditions and no infiltration topographic potential, sediment flow and net erosion/deposition (will be added later)

Based on field data (1m resolution)

• modified USLE LS
  
• USPED topo
  
• SIMWE topo (will be added later)

4. Comparison with the Cs data

Interpolated erosion/deposition pattern from Cs measurements

• ID for sampling points
  
• values of erosion/deposition estimated from Cs
  
• Cs measurements
  

Modeled erosion/deposition pattern compared with Cs measurements represented by point symbols (colored and sized according to the estimated erosion or deposition rate.

• USPED erosion/deposition with variable C
  
• USPED e/d, var. C - averaged (smoothed)
  

Maps for presentation

• raw 2.5m resolution IFSAR DEM with imagery draped over it, 4x vertical exaggerration - these are the elevation data used as a starting point
  
• DEM reinterpolated and smoothed to 2m resolution with contours and point IDs
  
• DEM with interpolated Cs erosion/deposition estimated draped as color and showed as size of the point symbols
  
• modified USLE and CS erosion/deposition as point symbols
  
• USPED erosion/deposition with variable C-fac, exponent=1.6, Cs as point symbols
  
• USPED erosion/deposition with variable C, exponent = 1 (this is just for comparison)
  
• Slope and CS estimates
  
• Water flow (Upslope area) and Cs estimates
  
• Simwe estimate of erosion/deposition with wild guess of parameters (that's why there is an erosion strip between the two soils instead of deposition) - I would need about 1 full day to look up the parameters. Anyway, this model actually shows deposition around points 15 and 14 (others don't). There are many points which seem to be shifted in the predcition only couple meters - the horizontal position of IFSARE data and samples may not be 100% accurate - that's why it would still be worth it to work with the field elevation data.