(505) 336-2600

Introduction

This fall we had the opportunity to work with Ted Miller, advised by Bruce Harrison, Ph. D and other colleagues at New Mexico Institue of Mining and Technology, New Mexico’s State Geologist, Matt Zimmerer, Ph.D., and others to support an undergraduate study on the benefits of using today’s drones and sensor technologies to possibly enhance techniques and methods into current soil mapping methods. Although the study is not complete, as of this date, it is well underway and the drones’ work is done.  Now Ted must satisfy the close scrutiny of the well-qualified advisors and professors that will ensure his scientific case remains valid. I don’t envy his task, but the greater benefits in life come from hard diligent work and he is doing that.  Best Regards and Wishes Ted, it was a pleasure!

Also, our gratitude goes to Federal Wildlife Officers of The Sevilleta National Wildlife Refuge a protected area of New Mexico managed by the United States Fish and Wildlife Service as part of the National Wildlife Refuge System.

Project Preliminary Summary by Ted Miller

Current soil mapping methods are time-consuming and expensive, especially at small scales and in remote areas. Traditional methods require substantial field work including soil pits, augers, and aerial photographs. Digital soil mapping can provide precise and accurate data on the soil surface and subsurface properties. Done using remote sensing algorithms that use spectral bands of data to calculate quantitative hydrological data, such as root zone soil moisture, that characterizes subsurface soil properties. In our study, the soil moisture is calculated using Surface Balance Algorithm for Land (SEBAL). The spectral data is collected using drone-based cameras; this increases both the spatial and temporal resolution of the product. The goal of this project is to use spectrally derived root zone soil moisture to create a soil map. No other study has used drone-based data for a SEBAL calculation.

We test the hypothesis that remotely sensed root zone moisture is directly related to soil properties. By stacking the soil moisture images from different days, we believe we can identify soil properties based on established soil drying curves. A single image informs us of the soil moisture on that day, but by stacking them, we can identify different soil moisture properties between soils. Field soil moisture measurements are taken to verify the precision of the method. The field site has been the subject of several other studies and is carefully mapped. We correlate the soil moisture properties we measure previous data such as silt, clay ratios, slope degree, organic matter content, and slope position to identify critical factors in soil moisture. We expect soil moisture controlled by slope aspect, clay content, and organic matter content. In this way, soil moisture can represent an approximation of soil properties.

Soil properties are used in many ways; by farmers and ranchers to optimize their agriculture and evaluate their land. Hydrologists use soil properties in computer modeling. Engineers use these properties for structural stability and planning. Unfortunately, the current soil maps are often inaccurate and outdated. Due to the traditional, labor-intensive method of making soil maps. Old methods require someone to dig holes and rely on coarse satellite imagery physically. However, times are changing. New digital methods are making data-driven maps that rely on multispectral imaging.

One method of obtaining soil moisture from spectral bands is the Surface Energy Balance Algorithm for Land (SEBAL), which uses remotely sensed TIR and NIR data in junction with meteoric data collected at weather stations to calculate Net Radiation, Soil Evaporative Flux, and the Evaporative fraction (Bastiaanssen, 1998). This method has been confirmed to be within 0.05 cm3/cm3 (by volume) in arid environments 90% of the time (Scull et al., 2003). Already, this method has been used to predict soil boundaries in New Mexico with success (Webster 1973, 1978; Engle, 2009). However, the current spatial resolution of this method is limited to 60m due to the satellite collection method. The temporal resolution is limited to the passing of the Landsat satellites, which is approximately every 6-8 days.

With the progress of technology, we now have the option of collecting spectral data using Unmanned Aerial Vehicles (UAVs), more commonly known as drones. UAVs can collect visible, NIR, and TIR spectral data over a large area in a short amount of time at a very high resolution (10-20 cm) Figure 1. UAVs have been used map plant health and moisture conditions for agricultural use with success (Kalantar, 2017; Hendrickson, 2000; Krishn

Figure 2. Taken from Bass et al. showing the location of the Sevilleta National Wildlife Refuge and study area

a, 2016). The proposed method uses drones and the SEBAL calculation to measure root depth soil moisture. These data correlate with existing soil property data.

 

Study Area

The study area is a first order drainage basin in the Sevilleta National Wildlife Refuge in central New Mexico (34˚24’ Lat 106˚59’ Lon) Figure 2. The basin is small with a total catchment of 0.034Km2. The soil parent material is Pleistocene age alluvial fan deposit sourced from the nearby Ladrones mountains. The alluvium comprised of sand, gravel, and boulders composed of schist and quartzite (McMahon, 1998). The climate in the area is semiarid with a mean annual temperature of ~15˚C and annual (mean)  precipitation of 255mm.

The Sevilleta Wildlife Refuge is currently housing the Long-Term Ecological Research (LTER) program which provides free daily data of wind speed, temperature, solar radiation, and precipitation. Two such weather stations located near the study area, station 43 to the north and station 45 to the south. Station 43 is within 1 km of the drainage basin and provides the temperature and wind speed data necessary for the SEBAL calculation.

The drainage is oriented east-west producing one north facing slope, one south facing slope. The north-south slope orientations lead to distinct vegetation contrast due to different amounts of water, evaporation, and sunlight (Gutierrez, 2004; McMahon, 1998). The north-facing slope covered by juniper trees with a ~6m rooting depth, the south slope is covered mainly by creosote bushes with a ~3m rooting depth. SEBAL uses NDVI to estimate subsurface soil moisture, the rooting depths for these plants represent the maximum extent of our soil moisture measurements. These differences in vegetation, sunlight, water, and other factors have led to the development of different soil depths and textures.

The south-facing slopes are characterized by thin A horizons, with little or no Bw or Bt horizons present and random distributions of carbonate accumulations (McMahon, 1998). The average depth for these soils averaged 5 cm. These south-facing slopes show strong calcium carbonate accumulation in the Ck horizon. Silt content decreases downslope with little variation in the clay content (McMahon, 1998). The north-facing slope soils are much deeper (~75 cm) have higher organic content, clearer soil boundaries, and more carbonate development relative to the south-facing slope, Figures 3 & 4. The soil layers are typically A 0-4 cm, Bt 4-18 cm, Bk 26-75 cm, Ck 75+ cm. The amount of clay and silt is much more significant on the north facing slopes, Figure 5, and varies depending on the catenary position between 9-12 g/cm2(McMahon, 1998). The amount of solar radiation also varies significantly between the slopes with the north facing slope experiencing ~20% less per year. Soils on the south face have higher runoff than the north face (McMahon, 1998). The north facing slopes also receive less sunlight than the south facing slopes. As expected the soils on the north facing slope retain more water and have lower evaporation than the south-facing slope, Figure 6 (Gutierrez, 2011). Resulting in less evapotranspiration on the north facing slope than the south (McMahon, 1998; Gutierrez, 2011). We expect to see this in our final map.

Theodore Miller

New Mexico Institute of Mining and Technology

Geology Masters Candidate

Figure 3. Taken from McMahon, 1998. Shows organic matter content by slope face

Figure 4. Taken from McMahon, 1998. Shows calcium carbonate content by slope face.

 

Figure 5. Taken from McMahon 1998. Shows silt and clay content by slope face.

Bibliography:

  1. Anderson, M. P., Woessner, W. W. & Hunt, R. J. Applied Groundwater Modeling: Simulation of Flow and Advective Transport. (Academic Press, 2015).
  2. Bass, B. Seasonal Shifts in Soil Moisture throughout a Semiarid Hillslope Ecotone during Drought: A Geoelectrical View. Soil Science Society of America Journal(2016).
  3. Carcione, J., Herman, G. & ten Kroode, A. Seismic modeling. GEOPHYSICS67,1304–1325 (2002).
  4. Cook, B. I., Bonan, G. B. & Levis, S. Soil Moisture Feedbacks to Precipitation in Southern Africa.J. Climate19,4198–4206 (2006).
  5. Dobos. Use of combined digital elevation model and satellite radiometric data for regional soil mapping. Available at:http://www.uni-miskolc.hu/~ecodobos/cikk.pdf.(Accessed: 26th October 2017)
  6. Engle, Emily. DIGITAL SOIL BOUNDARY DETECTION USING QUANTITATIVE HYDROLOGIC REMOTE SENSING. (2009).
  7. Hendrickson, L. L., Han, S. & Dickson, M. A. Method of determining and treating the health of a crop. (2003).
  8. Hengl, T. & Heuvelink, G. New challenges for predictive soil mapping. Global Workshop on Digital Soil Mapping. September 14-17, 2004(2004).
  9. Howell, D. et al.Chapter 34 Fitting Soil Property Spatial Distribution Models in the Mojave Desert for Digital Soil Mapping. n Developments in Soil Science(eds. Lagacherie, P., McBratney, A. B. & Voltz, M.) 31,465–624 (Elsevier, 2006).
  10. Kalantar, B., Mansor, S. B., Sameen, M. I., Pradhan, B. & Shafri, H. Z. M. Drone-based land-cover mapping using a fuzzy unordered rule induction algorithm integrated into object-based image analysis. International Journal of Remote Sensing38,2535–2556 (2017).
  11. Krishna, K. R. Push Button Agriculture : Robotics, Drones, Satellite-Guided Soil and Crop Management. (Apple Academic Press, 2016). oi:10.1201/b19940
  12. Lagacherie. An Introductory Perspective. Digital Soil Mapping31,1–620 (2007).
  13. Lagacherie, P. Digital Soil Mapping. 31,(El Sevier, 2006).
  14. McMahon, Dennis. Soil, Landscape and Vegetation Interactions in a Small Semi-arid Drainage Basin: Sevilleta National Wildlife Refuge, New Mexico. (1998).
  15. Morgan & Bui. Spatial data mining for enhanced soil map modeling. International Journal of Geographical Information Science16,533–549 (2002).
  16. Scull, P., Franklin, J., Chadwick, O. A. & McArthur, D. Predictive soil mapping: a review. Progress in Physical Geography: Earth and Environment27,171–197 (2003).
  17. Seneviratne, S. I., Lüthi, D., Litschi, M. & Schär, C. Land-atmosphere coupling and climate change in Europe. Nature443,205–209 (2006).
  18. SMJ Baban & Yusof. Mapping land use/cover distribution on a mountainous tropical island using remote sensing and GIS. International Journal of Remote Sensing1909–1918 (2001).
  19. Vivoni, E. R., Entekhabi, D., Bras, R. L. & Ivanov, V. Y. Controls on runoff generation and scale-dependence in a distributed hydrologic model. Hydrol. Earth Syst. Sci.11,1683–1701 (2007).
  20. Wierenga, Hendrickx & Nash. Variation of soil and vegetation with distance along a transect in the Chihuahuan desert. Journal of Arid Environments53–63 (1987).

 

 

Pin It on Pinterest