- State/s: CA and NV (proposed to add in AZ as well)
- MLRAs: 30 and 31
The original objective of this study was to determine the thermic/hyperthermic boundary across MLRA 30. Currently the primary objective is to estimate soil temperature regimes across the Mojave and Lower Colorado Desert. A secondary objective is to develop a long term record (> 10 years) for numerous sites within the region.
- Verification of current soil temperature regime estimates
- Long-term measurements to capture the spatial and temporal variability across MLRA 30 and 31
- Correlation of soil temperature, soil moisture, air temperature, precipitation, frost free days, and plant species presence and abundance
This climate study began in 1998 as part of a national study run by the National Lab and led by Henry Mount (Mount and Paetzold, 2002). The objective was to determine if the hyperthermic line was consistent across the southern US, which at the time was thought to be ~3000 in elevation. Beginning in 1998 five HOBOs were placed in Joshua Tree National Park, of which two were placed @ 5000 to determine if that elevation would be a realistic separation between thermic and mesic soil temperature regimes (STR). From 2000 to 2002, ~20 additional HOBO Stowaway sensors (placed at 50cm) and/or 4-channel loggers (placed at -1m, 0cm, 10cm and 50cm) were placed across Joshua Tree National Park (JTNP), the Chocolate Mountains, Fenner Valley, Mojave National Preserve, Death Valley National Park and Anza Borrego State Park. In 2003 and 2004, BLM OHV Area soil surveys in CA were in full swing, so another 25 HOBO loggers were placed across these three survey areas to help determine their STR. Four sites were also placed in the Mount Charleston Area at type locations by Leon Lato, Soil Scientist in Las Vegas, to establish concepts for high elevation mapping in the middle of MLRA 30. Another set of HOBOS (often referred to as the MDWC HOBOs) was set-up as a soil temperature catena for future work within the Mojave Desert West Central (MDWC) Soil Survey Area. The MDWC sites run from the top of the San Bernardino Mountains down to Melville Dry Lake (i.e. from frigid to hyperthermic was the hypothesis). These data were compiled and analyzed to: 1) determine where we needed specific information in our active survey areas and 2) to develop a tabular soil temperature model related to elevation and aspect (as a desktop reference for the office). This work occurred before GIS was common place.
In 2007 the soil survey of JTNP was restarted and the tabular model updated and 2 additional sites were place adjacent to JTNP.
In 2009, David Howell developed a digital soil map (DSM) of MAST with a spatial resolution of 60-meters using our ~ 8 years of soil temperature data in order to assist with the soil survey of JTNP. Davids results served as a guide for the remaining length of the JTNP Soil Survey. The model formula was:
MAST^2 (Celsius) = 5.72 (0.90 * Soil Adjusted Vegetation Index) 0.0009 * Elevation) + (0.011 * Slope) + (0.04 * Annual Solar Radiation) 0.10 * Summer Solar Radiation).
In 2011, work began on the soil survey for Mojave National Preserve (MNP). At that time Stephen Roecker updated the DSM of MAST developed by David Howell to expand its spatial extent and add missing sites. The updated model formula was:
MAST (Celsius) = 40.23 (0.009 * Elevation) (-0.23 * Coefficient of Variation in Annual Solar Radiation) (0.02 * Tasseled Cap Component 1).
This resulted in a more a stable model, and accounted for some discrepancies between the observed and predicted temperatures (i.e. trend in the residuals). However, analysis of this updated model stilled show wide standard errors for steep slopes, so eight additional sites were added to MNP with four being placed on opposing steep north and south slopes. In 2013, after one year of data was collected within MNP, the model was again updated.
MAST (Celsius) = 18.01 + (0.85 * PRISM Annual Air Temperature (Celsius)) (0.2 * Coefficient of Variation in Monthly Solar Radiation (%)) (0.01 * PRISM Annual Precipitation (mm)) (0.02 * Tasseled Cap Component 1)
Thanks to the additional sites placed on steep slopes their standard errors were somewhat reduced. Also by including precipitation as a variable this new model accounted for discrepancies between MAST and ecological sites above Anza Borrego that receive higher than normal precipitation. Secondly it was noted that by replacing elevation with air temperature this model had a more consistent soil/air temperature offset with data from weather stations.
In addition to our MAST modeling efforts there has also been two published studies on the Mojave. The first was by Schmidlin et al. (1983) who examined both the Great Basin and Mojave Deserts in Nevada. The second was by Bai et al. (2010) who examined the Mojave Desert in California. Both studies developed regression models using elevation, but Schmidlin et al. (1983) also incorporated latitude. For comparison, Stephen constructed raster layers of these models. The results showed that the model by Bai et al. (2010) displays considerably larger areas of hyperthermic soils than any of the other models. This made be due to the unconventional method used by Bai et al. (2010) to measure MAST. The model by Schmidlin et al. (1983) was very similar to the most recent iterations by Stephen. Since neither of the models by Bai or Schmidlin take into account aspect, Merkler et al. (2013) proposed that an aspect adjusted latitude be used. However, comparison of the spatial distribution of this model showed that the aspect adjustment was too strong. For example the adjustment produced extreme hyperthermic (> 25 oC) soils on south slopes, and mesic soils on the opposing north slopes.
The justification to continue the collection of soil temperature measurements across the Mojave and Lower Colorado Desert is to refine our existing model and collect a long term (> 10 years) record for numerous sites. In the past estimates of MAST have been subjective and have produced inconsistencies between published soil surveys. Therefore we need to establish a standard estimate of MAST to aid in the correlation with, and update of, published soil surveys. This same inconsistency exists with historic ecological sites. For example, within MLRA 30 and 31 we have commonly distinguish between cool (15-18 C) vs. warm (18-22 C) thermic areas as they support a different abundance and assemblage of vegetation. Therefore thanks to our MAST dataset and model we now have a better understanding of the relationship between MAST and vegetation.
Even though we have developed a respectable model, we still have work to do. An analysis of our present distribution of sites shows that we have 32 redundant sites and 3 areas that are underrepresented in our sample. Therefore we have developed a flexible sampling plan to retire our redundant sites and add 5 new ones. A preliminary analysis has shown that we could generate our present MAST model with 45 sites.
Soil series or other category (landscape unit for example) affected. Is the series a benchmark soil? The landscape unit affected would be MLRA 30 and 31. Our intent has never been to look at a particular benchmark soil, but to try to develop an overall MAST model of our MLRAs, especially MLRA 30. For example, in JTNP there were areas that didnt agree with our MAST concepts from our work in 2000 and 2001 when it was restarted in 2007. The longer term soil temperature data for JTNP, combined with extensive ecological site field work, convinced us to change soils in the upper part of Pinto Basin from hyperthermic to thermic.
Now that we have 13+ years of data for many of our MAST sites, our model for the entire MLRA suggests that there are many soil series as well as map units developed for CA697, Fort Irwin Soil Survey Area, and CA699, Marine Core Air Ground Combat Center in Twentynine Palms, CA, that need the soil temperature regime updated and map units changed. In Fort Irwin, many of the soils on the southern end are warm thermic soils and were mapped across the whole thermic temperature regime in the mid-90s. We now realize that our ecological sites correlate very easily if we differentiate our soil series concepts as warm thermic versus cool thermic. An example of a soil series that needs to be warm thermic is the Stonegold series.
For CA699, more than just soil series will need to be updated for soil temperature. When that survey was mapped in the mid-90s, the idea that was brought over from NV was that the hyperthermic line was at 2000. This meant that several map units were mapped as thermic because they were above 2000 in elevation, but now that we have actual soil temperature data near this older survey area, we know that soils are hyperthermic up to at least 2600 on flat areas and even higher (closer to 4000) on south-facing aspects on hills and mountains. There will need to be update work on OSD type locations for a few soils (e.g., Narea and Desfirex soils) and map unit concepts for the map units that include these named components. Update projects for the work in CA699 as well as CA697 will be created in NASIS.
Ecological site concepts and Ecological Site Descriptions are in development. The data collected during this project will enhance the quality of the ongoing work.
We have seen clear correlations in Joshua Tree National Park and CA698, Mojave Desert, West Central Area, between ecological sites and STR. By continuing to collect soil temperature data for the long term in the Mojave and Lower Colorado Deserts we can continue to tighten up our ecological site concepts.
Key soil information includes soil temperature, air temperature, precipitation, soil moisture, and site information including pedon descriptions . Site information will be stored in NASIS.
See Site and Sensor Table, attached. Fill out the Site and Sensor Table for all sites you want to include in an official record. In this section, describe office procedures used to plan, log, and collect currently monitored soil climate sensor sites. For sites currently being monitored, create a summary table with the following structure and paste it here:
Current soil climate data monitoring sites, examples in italics
mlra_sso user_site_id mlra Temp regime Moist regime Purpose of sensor arcata XXXXXXXXXX 4B isomesic udic long term data victorville XXXXXXXXXX 31 ? aridic hyperthermic vs. thermic
Support materials available to support project implementation and data interpretation: List current data and literature that inform the current soil temperature and moisture regime estimates and that are assisting with verification, update, and regional modeling. Include current models. Add more as needed.
- National Elevation Dataset (NED)
- Parameter-elevation Regressions on Independent Slopes Model (PRISM)
- Landsat tasseled cap components
- NAIP imagery (current and historic)
- Current and historical soil surveys
- Ecological site data
- Bai, Y., T.A. Scott, W. Chen, R.C. Graham, L. Wu, A.C. Chang, and L.J. Lund, 2010. Soil Temperature Regimes in the Mojave Desert. Soil Science, 175(8):398-404.
- Howell, D., 2009. What does it take to create a digital soil mapping model, and what are all those cryptically named raster on the computer? pp.7.
- Laity, J., 2008. Deserts and Desert Environments. Wiley-Blackwell, West Sussex, UK.
- Merkler, D.J., B.H. Gore, and K. Holcomb, 2013. A geospatial model of soil temperature for the Basin and Range. Poster presentation at the 2013 Soil Science Society of America Annual Meetings.
- Mount, H.R., and R.F. Paetzold, 2002. The temperature regime for selected soils in the United States. United States Department of Agriculture, Natural Resources Con-servation Service, National Soil Survey Center, Lincoln, Nebraska, Soil Survey Investi-gation Report No. 48.
- Peterson, F.F., 1992. Status of Soil Climate Studies in Nevada. pp1-11.
- Roecker, S.M. and C.A. Haydu-Houdeshell, 2012. Modeling and Application of Soil Temperature in the Mojave and Lower Colorado Deserts of California. 2012 Western Regional Cooperative Soil Survey Conference.
- Schmidlin, T.W., F.F. Peterson, and R.O. Gifford, 1983. Soil Temperature Regimes of Nevada. Soil So. Sci. Am. J., 47:977-982.
See start and end dates in Site and Sensor Table, attached.
Presently we have 77 active MAST sites, and 12 SCAN sites throughout MLRA 30 and 31. In the near future we intend to retire approximately 32 redundant sites, and install 5 new sites in underrepresented areas, for a total combination of 45 HOBO and SCAN sites.