Bioclimatic and soil moisture monitoring across elevation in a mountain

watershed: Opportunities for research and resource management

E. C. Osenga1, James C. Arnott1, 2, K. Arthur Endsley2, and J. W. Katzenberger1

1
Aspen Global Change Institute.
2
University of Michigan, School for Environment & Sustainability.

Corresponding author: Elise C. Osenga1 ([email protected])

Key Points:
 Soil moisture is key to understanding and predicting change in hydrology and ecology
amid climate variability and change
 In situ soil moisture and weather monitoring data are now available across an 1800 m
elevation span in a mountain watershed
 The network is supported and guided by resource managers and supports both
research and resource management goals

This article has been accepted for publication and undergone full peer review but has not
been through the copyediting, typesetting, pagination and proofreading process which may
lead to differences between this version and the Version of Record. Please cite this article as
doi: 10.1029/2018WR023653

© 2019 American Geophysical Union. All rights reserved.

Abstract
Soil moisture data are critical to understanding biophysical and societal impacts of climate
change. However, soil moisture data availability is limited due to sparse in situ monitoring,
particularly in mountain regions. Here we present methods, specifications, and initial results
from the interactive Roaring Fork Observation Network (iRON), a soil, weather, and
ecological monitoring system in the Southern Rocky Mountains of Colorado. Initiated in
2012, the network is currently comprised of nine stations, distributed in elevation from 1,890
to 3,680 meters, that continually collect and transmit measurements of soil moisture at three
depths (5 cm, 20 cm, 50 cm), soil temperature (20 cm), and meteorological conditions. Time-
lapse cameras for phenological observations, snow depth sensors, and periodic co-located
vegetation surveys complement selected stations. iRON was conceived and designed with the
joint purpose of supporting bioclimatic research and resource management objectives in a
snow-dominated watershed. In the short term, iRON data can be applied to assessing the
impact of temperature and precipitation on seasonal soil moisture conditions and trends. As
more data are collected over time, iRON will help improve understanding of climate-driven
changes to soil, vegetation, and hydrologic conditions. In presenting this network and its
initial data, we hope that the network’s elevational gradient will contribute to bioclimatic
mountain research, while active collaboration with partners in resource management may
provide a model for science-practice interaction in support of long-term monitoring.

Plain Language Summary

As climate change drives shifts in temperature and precipitation, researchers and resource
managers can benefit from improved monitoring of soil moisture. Understanding the
relationship between soil moisture and other system components is crucial to improving
water availability projections and understanding ecosystem responses to climate change.
Despite their significance, in-ground soil-moisture measurements are often not available
across multiple elevations within a single watershed.

This paper presents a network in the Southern Rocky Mountains intended to help address this
data gap and compliment data from other networks. The interactive Roaring Fork
Observation Network (iRON) consists of 9 locations across an 1,800 m change in elevation.
Each station measures soil moisture at three depths, soil temperature, air temperature,
humidity, and precipitation. Some stations are equipped with cameras or snow depth gauges,
and for 8 sites vegetation surveys are conducted. The data are available through a simple data
portal.

The network was established with local resource manager support, and one of its guiding
purposes is to support management and restoration planning efforts. Because of the network’s
on-going monitoring across multiple elevations and habitats, iRON will provide researchers
and resource managers with access to valuable information about changes in soil conditions
in a changing climate.

© 2019 American Geophysical Union. All rights reserved.

1 Introduction

Soil moisture dynamics are critical to characterizing regional climate change impacts
on hydrology and ecosystems. Although the full extent of soil moisture and climate
interactions is a developing area of research (Seneviratne et al., 2010), minimum near-surface
soil moisture is projected to decrease in the southwestern United States as climate change
continues (Whener et al., 2017). Soil moisture is increasingly understood to be a key driver—
and indicator—of regional hydrologic variability and change (Seneviratne et al., 2010), and
climate-driven alterations to soil moisture have repercussions for both ecological health (Pecl
et al., 2017; Robinson et al., 2008) and human well-being (Lempert & Groves, 2010). Despite
the importance of soil moisture in understanding hydrologic systems, limited observational
data have hampered both understanding of the relationship between soil moisture and runoff
and the ability to develop and validate hydrologic models. Better understanding of soil-
moisture dynamics has the potential to advance research and better support resource
management in the context of climate change (Seneviratne et al., 2012).

In the southwestern United States, water managers face climate-driven disruptions to

water supplies (Barnett et al., 2008) and, simultaneously, challenges to meeting the needs of
rapidly growing populations (Dettinger et al., 2015). As future hydrologic conditions will
likely depart from historical patterns, models and forecasts of streamflow are becoming
increasingly relevant to water management for both near-term, e.g., winter and early season
(Pagano, 2010), and long-term planning, e.g., decadal to centennial (Udall & Overpeck,
2017). However, current models of runoff and other hydrologic processes in complex
terrains, such as mountain landscapes, often do not represent soil moisture well (Pagano,
2010).

Recent developments in remote sensing of soil moisture, particularly the launch of the
Soil Moisture Active Passive satellite in 2015 (Colliander et al., 2017; Entekhabi et al.,
2010), provide the opportunity for monitoring and modeling soil moisture across multiple
scales (Peng et al., 2017), but the resolution of remotely-sensed soil moisture data is often not
fine enough for watershed-scale applications in mountainous regions where topography and
soil moisture are heterogeneous, and remotely sensed soil moisture is often limited to shallow
depths (Cowley et al., 2017; Dobriyal et al., 2012). There are also a variety of on-the-ground
networks across the globe that include soil moisture (International Soil Moisture Network,
n.d.) including networks in the southwestern United States, such as NEON, SCAN, and select
sites in the SNOTEL program. However, because of the heterogeneity of climate and soil
moisture within mountain watersheds, additional data in previously unmonitored watersheds
has a potential to be valuable in augmenting existing in situ data and complementing
remotely sensed data. With mountains providing the headwaters for millions of water users, it
will be increasingly important to monitor soil moisture in the context of understanding water
availability (Lempert & Groves, 2010) and improving water forecasts for utility managers
(Pagano, 2010).

In addition to water resource management, soil moisture is also pertinent to ecosystem

health and ecosystem management in semi-arid climates (Whener et al., 2017). Along with
variables like vapor pressure deficit and precipitation, soil moisture is understood to be a key
factor in tree survival (Anderegg et al., 2015; Daubenmire, 1968; Worrall et al., 2010) and
fire risk (Bourgeau‐Chavez et al., 2007; Lavell et al., 2012; Williams et al., 2012).

© 2019 American Geophysical Union. All rights reserved.

 In light of the importance of soil moisture observations to emerging research and
management questions, we report here on the development of an in situ soil, meteorological,
and ecological monitoring network in the Southern Rocky Mountains of Colorado that
includes soil moisture measurements at 5, 20, and 50 cm depths. This network joins the
growing soil moisture research community and contributes a dataset distinguished by its
inclusion of in situ observations across multiple elevations in a single watershed. The
intention of this network is not only to serve as a monitoring project supporting local research
and resource management, but also to augment existing datasets, enabling researchers to
answer broader questions about climate impacts on mountain hydrology and ecology. Co-
designed with local land managers and other stakeholders, this network seeks to support both
scientific research and management needs. The purpose of this article is to describe the
context, monitoring set up, specifications, data, and data access for a broad audience of
potential data-users.

2 Methods and Context: Introducing the interactive Roaring Fork Observation

Network

2.1 Network overview & context

The interactive Roaring Fork Observation Network (iRON) is a series of in situ soil,
meteorological, and ecological monitoring stations. iRON is hosted and maintained by the
Aspen Global Change Institute (AGCI), a Colorado-based non-profit research organization
that works to advance understanding of global change. Its stations are situated across an
elevational gradient from 1,890 m (near a confluence with the Colorado River) to 3,680 m
(near the Continental divide at Independence Pass) in the Roaring Fork Watershed of the
Southern Rocky Mountains of Colorado (Figure 1). The Roaring Fork Watershed has an area
of 3,760 km2 and is a major tributary of the Upper Colorado River Basin. As a headwaters of
the Upper Colorado, the Roaring Fork River’s flows are critical to meeting present and future
water demands of the western states of the Colorado River Compact, as well as the
downstream water demands of Mexico. In aggregate, the Colorado River serves around 40
million people (Bureau of Reclamation, 2012). Numerous studies of the Upper Colorado
Basin indicate that climate change will reduce streamflow in the coming decades, affecting
recharge of the major reservoirs of Powell and Mead and increasing the likelihood of supply
shortages (Castle et al., 2014; Dettinger et al., 2015; Udall & Overpeck, 2017; Vano et al.,
2013). Abundant concern about climate-related risks to land and water resources, together
with the watershed’s significance to downstream communities, provides compelling reasons
to conduct long-term observations of soil moisture and other environmental variables.

2.2 Network co-design with resource managers

A key feature of the development of iRON has been intensive collaboration with local
resource managers to co-design and help sustain the network. Prior to its establishment in
2012, local interest in bioclimatic change in the Roaring Fork Valley was documented
through studies that surveyed and interviewed practitioners working to manage and conserve
water and land resources (Arnott et al., 2014; Arnott et al., 2015; Aspen Global Change
Institute, 2006). This process occurred through multi-stakeholder roundtables focused on
water and forestry issues, as well as through targeted interaction with specific land
management entities. Local partnerships have ranged from financial support for network
establishment and maintenance to identification of opportunities where monitoring stations
can support planning, restoration, and evaluation for adaptive management. An example of

© 2019 American Geophysical Union. All rights reserved.

potential data application includes comparison of changes in evolving bioclimatic conditions
with species-specific tolerances to guide restoration decisions. During the establishment of
the network, input from management entities was complemented by guidance from the
scientific advisory group, acknowledged at the end of this article. The iRON’s science
advisors provided input on network design and connections between potential research
agendas and stakeholder-relevant data applications.

The multi-use approach of iRON is facilitated by a public website which provides the
availability to view and access live and archived data (agci.org/iron), including access to an
automated data storage platform.

2.3 Site section, equipment, and data protocols

The primary criteria for iRON station selection has been distribution of monitoring
locations across the watershed’s elevational spread and main ecozones (Figure 2), including
shrublands, montane, and alpine environments with the addition of a sub-alpine site planned
for the near future. These ecosystems were subjectively selected because they represent
dominant ecosystem types within the watershed and are of particular ecological interest to
public land managers. Additional criteria used to determine the research sites were: land use
permissions for long-term placement and local management input. For example, two stations
were specifically selected to support pre- and post- monitoring of restoration treatments
planned by local land managers of a formerly grazed open-space and an impacted riparian
meadow.

Each station is equipped with set of dielectric soil moisture sensors, with one sensor at
each depth of 5, 20, and 50 cm. Additional equipment includes a soil temperature sensor at 20
cm and additional basic meteorological equipment mounted on a 2 m or 3 m tower (Table 1).
Two stations are additionally equipped with a Judd snow depth sensor, and one station
includes a time-lapse camera that takes time-stamped photographs every morning and
evening. Possible applications of the photographs are still being developed and include: the
potential to compare snow depth readings to images that may reveal patchiness in snow
persistence and opportunities for identifying phenological events such as flower blooms.

Two stations, Brush Creek and Spring Valley have been equipped with a second set of
soil moisture sensors. At Brush Creek, the duplicate set is being used to establish baseline
comparisons of a location that will be used as a control and a location that will be replanted
during county restoration efforts. In the case of Spring Valley, the second set of soil moisture
sensors are located approximately 3 m from the primary set of soil moisture sensors and were
included to allow for potential manipulative comparison experiments by local students, as the
station is located near to a Colorado Mountain College campus.

A new station at Castle Creek, slated for addition to the network in spring of 2019,
will expand on the standard instrumentation of iRON stations to include energy balance
measurements, as well snow depth and wind speed and direction. Opportunities for working
with relevant data from other networks are also being explored, including the limited LiDAR
data (Colorado Geological Survey, n.d.) available for the Roaring Fork Watershed and
consideration of data from NRCS SNOTEL sites, particularly the Schofield Pass site-- the
only SNOTEL station in the watershed that includes soil moisture (Natural Resource
Conservation Service, n.d.).

© 2019 American Geophysical Union. All rights reserved.

 Data are collected every 20 minutes at cellular data transmission stations and every
hour at one satellite-uplink station. All loggers transmit their data to an online server every 4
to 6 hours. In addition to continuous data collection by the equipment at each site, vegetation
surveys are conducted to track species presence and abundance over time. Modified Whitaker
Plot surveys (Stohlgren et al., 1995) are conducted at each site on a rotating 3-year basis.
Initial Modified Whitaker Plot surveys have already been completed for 8 of the site
locations. Additional tree-specific surveys are planned for each site. Routine equipment
maintenance is performed annually, with additional site visits as-needed for instrumentation
repair. An alarm is set to trigger for potentially false readings from equipment.

Gravimetric calibrations were carried out for all soil moisture instrumentation:
Decagon EC-5 dielectric sensors (for 5 cm readings) and 10HS dielectric sensors (for 20 cm
and deeper readings). The EC-5 and 10HS sensors were calibrated in-lab by taking sensor
readings of soil moisture after recorded volumes water were mixed into a known volume of
soil collected from each site. A regression equation was developed for each depth and station
to relate sensor readings to actual soil moisture volumes. Root mean squared error (RMSE)
was calculated based on the observed soil moisture volume and soil moisture predicted by the
regression equations. RMSE ranged from 0.010 m3/m3 (at Northstar Transition Zone, 20-cm
depth) to 0.087 m3/m3 (Northstar Transition Zone, 50-cm depth) with a median RMSE of
0.027 m3/m3. Measurement accuracy at some sites may have been impacted by soil texture
and mineral composition. A full table of RSME values by station and soil depth can be found
in supporting information materials (Table S2). In the available literature, calibration results
for the Decagon sensors were within +/– 0.02 m3/m3 to 0.05 m3/m3 accuracy of soil moisture
for most soil mineral compositions in laboratory settings (Kizito et al., 2008). In-lab
calibrations for iRON yielded similar results (Osenga, 2018a).

Other station instrumentation was tested for functionality in-lab but additional, site-
specific calibration was not carried out. Manufacturer standards for equipment accuracy can
be found in the supporting information materials (Table S1).

2.4 Data management and accessibility

Real-time data are telemetered from iRON stations every 4 hours to Hobolink, a
cloud-based system for storing and accessing remote monitoring data, operated by the Onset
Computer Corporation. The raw data are then delivered to AGCI’s server by secure file
transfer protocol (SFTP) and are sorted, stored, flagged, and made available to users through
an application programming interface (API). An API is any set of tools and protocols that
enable other software to be built; scientific APIs are intended to allow new or existing
software to connect to some resource or information, typically through an internet
connection.

iRON’s API is described and hosted on the iRON Data Board (irondataboard.org).
This interface allows for customizable data exports by range of time and variables. Currently,
data are available from December 2017 forward on this site, and users can filter by time and
station. Data delivery to the server and the server’s internal consolidation of new data occur
once every 6 hours. The iRON Data Board automatically flags values (using the “valid” field,
with values “Y[es]” or “N[o]”) that are out of range for a given measurement type. In general,
any additional rules can be added, facilitating automated quality assurance and quality
checking (QA/QC). These rules, like the soil calibration equations, are applied on top of the

© 2019 American Geophysical Union. All rights reserved.

raw data, ensuring that raw data are never changed and providing the flexibility for future
added value, such as the development of more sophisticated soil calibration.

The iRON Data Board includes a browser-based API that uses URL strings to form
requests for data with a standards-based design advantageous for the representation and
serialization of geophysical data (Endsley & Billmire, 2010). A small number of assets are
hosted by the iRON Data Board, corresponding to endpoints in the browser-based API. Most
users will be interested in the “Readings,” which are actual measured values for a given
station. The “Readings” endpoint accepts a few parameters, such as the “from” and “to”
parameters required to specify the date and time range for which observations are requested.
The latency, or time delay, between making a data request and the initialization of a
download increases with the size of the data request, and API requests for more than 90 days
of data across one or more stations will be denied by the server. Therefore, it is recommended
to that API requests be made for individual station data by specifying the “station_id” (see
Table 1) or a portion of the station name with the “station” parameter. If longer than 90 days
of data are needed, exports of archived data through 2018 are available on the iRON Data
Board website (irondataboard.org).

Additional functionality has been developed to include the ability to filter for specific
variables, such as requests for either calibrated or uncalibrated soil moisture data (calibrated
data are the default), the measurement units used (metric or imperial), and the time zone of
time stamps. In addition to rich interfaces for accessing iRON data, the iRON Data Board
provides internal quality assurance and periodic data backups. Archived public datasets are
additionally searchable on Zenodo.org via an ORCID identifier (ORCID 0000-0002-2747-
2994).

3 Initial Results & Discussion

Data records for iRON stations currently range from 2.5 – 6.5 years. While the data
record is insufficient in length to characterize trends at this time, the existing observations
demonstrate the network’s potential for long-term research, as well as its near-term utility
(Figure 3). Existing data are already being used to characterize wetting and drying events on
multiple temporal scales and to provide comparisons across the elevational gradient. The
frequency of data collection (every 20 minutes or every hour) allows for observation of soil
wetting events on short temporal scales—e.g., tracking penetration of precipitation across 5,
20, and 50 cm soil depths over the course of hours to days. Figure 4a provides an example of
how the frequent collection of data reveals dynamic soil responses to rain events. In this
example, rain events over the course of the 7-day period totaled 3.2 cm (Figure 4a). Because
of the frequency of data collection, it was possible to see the time lag between when this
wetting event penetrated to a 5 cm depth and when it penetrated to a 20 cm depth—within the
same day. For this event, moisture did not penetrate to a 50 cm depth.

Initial results also show the impact of seasonal events on soil moisture throughout the
growing season, particularly highlighting the role snowmelt in early spring as critical in
increasing soil moisture prior to late spring and summer drying periods. The Roaring Fork
Valley is a snow-dominated watershed, and the significance of the snowpack on soil moisture
can be seen in soil moisture response, where all depths (5, 20, and 50 cm) show recharge that
brings soil moisture near to saturation in early spring during ground thaw and snowmelt. In
this 2013 example, rain events of less than 0.5 cm occurring during the summer season are
insufficient to increase soil moisture at a 20 cm depth, and soil moisture at a 50 cm increased
only slightly after even the largest summer rain events of 1.5 cm or more in a single day

© 2019 American Geophysical Union. All rights reserved.

(Figure 4b). Across multiple sites and multiple wetting events, soil moisture at a 5 cm depth
was commonly found to be more variable than soil moistures at greater depths

The applicability of the network in addressing bioclimatic questions is further

augmented by its geographic scale, which spans much of the elevational gradient of the
watershed. Observations from iRON provide a basis for tracking and comparing future
changes in timing of snowmelt and other hydrologic events at different elevations, such as the
date of soil saturation compared between the lowest-elevation and highest-elevation sites
(Figure 5).

As timing of snowmelt is anticipated to shift earlier with warming climates (Clow,

2010; Gillan, et al., 2010), data records that observe snow and soil moisture for multiple
elevations within a single climatic region may be used to identify elevational differences in
intensity of response to regional warming. Specifically, the difference in the timing of the
spring melt event across elevations and across years may be directly observed in the data.
Additionally, the correspondence between events such as soil saturation and snowmelt
indicates a potential to combine iRON data with datasets from outside the network to
contribute to hydrologic models and generate improved forecasts of events such as the timing
of snowmelt, runoff, and streamflow dynamics (Harpold et al., 2017; Mahanama et al., 2012).
Partnerships are currently being developed with researchers working on water models to
explore the possibility of using the Roaring Fork Watershed as a case study for applying
observational soil moisture data to improve the representation of soil moisture in hydrologic
models in mountainous terrain.

Understanding the long-term impacts of climate change for natural resources was a
primary motivation for local stakeholders engaged with the initial project development, and
ongoing conversations with community partners have been critical to ensuring local
relevance of and support for the network. In addition to their near-term utility, data from
iRON are also intended to reveal trends over time at longer-term scales (e.g. decadal),
including insight into ecological response to climate change. Data from the Modified
Whitaker plots have the potential to reveal changes in plant abundance or species type and
elevational migration by vegetation over time concurrent with trends revealed in the soil
moisture and meteorological data. Improved understanding of the role played by different
climatic and hydrologic mechanisms in vegetation invasion or mortality will be important in
determining future species ranges and vulnerability to climate change (Allen et al., 2015;
CNHP, 2015; Parida & Buermann, 2014), with application opportunities for land managers
and other stakeholders. Although identifying species shifts is a multi-decadal undertaking,
this project seeks to establish, at the least, baseline ecological records against which future
studies may be compared. Initial results from iRON reveal its potential for application in
understanding these ecology-climate-soil relationships. Moving forward, establishing
partnerships for further application of these data to regional and national scale research will
be critical, and it is hoped that such partnerships will aid in securing additional support for
this research through federal and local research grants.

4 Conclusion

The interactive Roaring Fork Observation Network can help both researchers and
resource managers to better understand the role of soil moisture in mountain watershed
ecology and hydrology. As noted by the Intergovernmental Panel on Climate Change (IPCC)
and abundant other literature, soil moisture is critically important to Earth systems research,

© 2019 American Geophysical Union. All rights reserved.

despite scarce in situ monitoring (Mahanama et al., 2012; Seneviratne et al., 2012; Whener et
al., 2017). Although satellite-based measurements of surface soil moisture have improved in
recent years, challenges remain in measuring variation in soil moisture across complex
terrains such as mountain ecosystems (Cowley et al., 2017). Filling gaps in existing soil
moisture monitoring networks will improve capacity to model the changing waterscape of
mountain regions and allow for more informed ecological and water management decisions
regarding mitigation and adaptation to climate change impacts.

Data collected through iRON can support both regionally-focused and more general
studies on ecological, climatological, and hydrological response to climate change and
variability in mountain areas. In addition, the network also provides a live, simultaneous
comparison of weather events across a mountain watershed. Examples of research pursuits
that could benefit from the incorporation and use of iRON data include:

 change in vegetation and soil moisture over time, including opportunities to validate
and inform models of climate-driven vegetation shifts;
 partitioning of precipitation into the atmosphere, soil reservoirs, and run-off;
 water availability forecasting, with a focus on modeling snowpack-to-streamflow
hydrologic dynamics;
 and the relationship between remotely sensed representation of soil moisture and in
situ observations across an elevational gradient.

AGCI continues to expand collaboration and encourage researchers working on regional

hydrologic or ecologic responses to climate change to join this effort. Additionally, the
design of this network as a partnership between local land management, researchers, and an
organization spanning the boundary between research and practice may serve as a useful
model for supporting the development, maintenance, use, application, and engagement of
bioclimatic monitoring elsewhere. We hope the iRON’s wide elevational gradient in the
Southern Rocky Mountains and its watershed-scale measurements can contribute to a better
understanding of the systems that are critical to sustaining mountain communities, including
our own.

Acknowledgments, Samples, and Data

AGCI wishes to acknowledge and thank the advisors who helped develop the initial concept
for this network, as well as those who provided input and assistance during its development:
Jeffrey Deems, Linda Joyce, David Lawrence, Delia Malone, Gerald Meehl, Jeffrey
Morisette, Michael Ryan, Todd Sanford, Michael SanClements, David Schimel, Diana Six,
Jeffery Taylor, and Alan Townsend.

We do not have any known conflicts of interest associated with publishing this report, nor are
there any conflicts of interest associated with the financial support provided for this research
or its outcomes.

Archived data used for this paper’s results and figures and supporting metadata can be found
at Zenodo.org, under the title “iRON Soil Moisture Calibrated 2013 to 2018,” DOI:

© 2019 American Geophysical Union. All rights reserved.

10.5281/zenodo.1271667 (Osenga, 2018b). The project API, the iRON Data Board can be
found at iRONDataBoard.org.

References
Allen, C. D., Breshears, D. D., & McDowell, N. G. (2015). On underestimation of global
vulnerability to tree mortality and forest die-off from hotter drought in the
Anthropocene. Ecosphere, 6(8), 1–55. doi:10.1890/ES15-00203.1
Anderegg, W. R. L., Flint, A., Huang, C. Y., Flint, L., Berry, J. A., Davis, F. W., … Field, C.
B. (2015). Tree mortality predicted from drought-induced vascular damage. Nature
Geoscience, 8(5), 367–371. doi:10.1038/ngeo2400
Barnett, T. P., Pierce, D. W., Hidalgo, H. G., Bonfils, C., Santer, B. D., Das, T., … Dettinger,
M. D. (2008). Human-induced changes in the hydrology of the Western United States.
Science, 319(5866), 1080–1083. doi:10.1126/science.1152538
Bourgeau‐Chavez, L. L., Kasischke, E. S., Riordan, K., Brunzell, S., Nolan, M., Hyer, E., …
Ames, S. (2007). Remote monitoring of spatial and temporal surface soil moisture in fire
disturbed boreal forest ecosystems with ERS SAR imagery. International Journal of
Remote Sensing, 28(10), 2133–2162. doi:10.1080/01431160600976061
Castle, S., Thomas, B., Reager, J., Rodell, M., Swenson, S., & Famiglietti, J. (2014).
Groundwater depleation during drought threatens future water security of the Colorado
River Basin. Geophysical Research Letters, 10, 5904–5911.
doi:10.1002/2014GL061055.Received
Clow, D. W. (2010). Changes in the timing of snowmelt and streamflow in Colorado: A
response to recent warming. Journal of Climate, 23(9), 2293–2306.
doi:10.1175/2009JCLI2951.1
Colliander, A., Jackson, T. J., Bindlish, R., Chan, S., Das, N., Kim, S. B., … Yueh, S. (2017).
Validation of SMAP surface soil moisture products with core validation sites. Remote
Sensing of Environment. doi:10.1016/j.rse.2017.01.021
Colorado Geologic Survey. (n.d.). LiDAR in Colorado. Retreived from
http://coloradogeologicalsurvey.org/geologic-mapping/lidar/

Colorado Natural Heritage Program [CNHP]. 2015. Climate Change Vulnerability

Assessment for, Colorado Bureau of Land Management. K. Decker, L. Grunau, J.
Handwerk, and J. Siemers, editors., & Colorado Natural Heritage Program, Colorado
State University, Fort Collins, C. (2015). Climate Change Vulnerability in Colorado,
(December), alpine 109-114. Retrieved from
http://www.cnhp.colostate.edu/download/documents/2015/CCVA_for_Colorado_BLM_
final.pdf
Colorado River Basin Water and Demand Study: Executive Summary. (2012). Retrieved from
https://www.usbr.gov/watersmart//bsp/docs/finalreport/ColoradoRiver/CRBS_Executive
_Summary_FINAL.pdf
Cowley, G. S., Niemann, J. D., Green, T. R., Seyfried, M. S., Jones, A. S., & Grazaitis, P. J.
(2017). Impacts of precipitation and potential evapotranspiration patterns on
downscaling soil moisture in regions with large topographic relief. Water Resources
Research, 53(2), 1553–1574. doi:10.1002/2016WR019907
Daubenmire, R. F. (1968). Soil moisture in relation to vegetation distribution in the

© 2019 American Geophysical Union. All rights reserved.

 mountains of Northern Idaho. Ecology, 49(3). doi:10.2307/1934109
Dettinger, M. D., Udall, B., & Georgakakos, A. (2015). Western water and climate change.
Ecological Applications, 25(1), 2069–2093. doi:10.1890/15-0938.1
Dobriyal, P., Qureshi, A., Badola, R., & Hussain, S. A. (2012). A review of the methods
available for estimating soil moisture and its implications for water resource
management. Journal of Hydrology, 458–459(January 2018), 110–117.
doi:10.1016/j.jhydrol.2012.06.021
Endsley, K. A., & Billmire, M. G. (2010). Distributed visualizations of gridded geophysical
data: the Carbon Data Explorer, version 0.2.3. Geoscientific Model Development, 9,
383–392. doi:10.5194/gmd-9-383-2016
Entekhabi, D., Njoku, E. G., O’Neill, P. E., Kellogg, K. H., Crow, W. T., Edelstein, W. N.,
… Van Zyl, J. (2010). The Soil Moisture Active Passive (SMAP) Mission. Proceedings
of the IEEE, 98(5).
Gillan, B. J., Harper, J. T., & Moore, J. N. (2010). Timing of present and future snowmelt
from high elevations in northwest Montana. Water Resources Research, 46(1), 1–13.
doi:10.1029/2009WR007861
Harpold, A. A., Sutcliffe, K., Clayton, J., Goodbody, A., & Vazquez, S. (2017). Does
Including Soil Moisture Observations Improve Operational Streamflow Forecasts in
Snow-Dominated Watersheds? Journal of the American Water Resources Association,
53(1), 179–196. doi:10.1111/1752-1688.12490
International Soil Moisture Network. (n.d.). Retreived from https://ismn.geo.tuwien.ac.at/
Katzenberger, J., & Masone, M. (2009). Colorado’s Roaring Fork River: Incorporating
climate change projections into watershed management. Mountain Research Initiative
News.
Kizito, F., Campbell, C. S., Campbell, G. S., Cobos, D. R., Teare, B. L., Carter, B., &
Hopmans, J. W. (2008). Frequency, electrical conductivity and temperature analysis of a
low-cost capacitance soil moisture sensor. Journal of Hydrology, 352(3–4), 367–378.
doi:10.1016/j.jhydrol.2008.01.021
Lavell, A., Oppenheimer, M., Diop, C., Hess, J., Lempert, R., Li, J., … Myeong, S. (2012).
Disaster risk, exposure, vulnerability, and resilience. In: Managing the Risks of Extreme
Events and Disasters to Advance Climate Change Adaptation. Cambridge, UK and New
York, NY, USA.
Lempert, R. J., & Groves, D. G. (2010). Identifying and evaluating robust adaptive policy
responses to climate change for water management agencies in the American west.
Technological Forecasting and Social Change, 77(6), 960–974.
doi:10.1016/j.techfore.2010.04.007
Mahanama, S., Livneh, B., Koster, R., Lettenmaier, D., & Reichle, R. (2012). Soil moisture,
snow, and seasonal streamflow forecasts in the United States. Journal of
Hydrometeorology, 13(1), 189–203. doi:10.1175/JMH-D-11-046.1
Osenga, E. C. (2018a). iRON Methods Used for Calibration of Decagon 10HS, EC5 Soil
Moisture Probes (Version 1). Zenodo. Retreived from
https://doi.org/10.5281/zenodo.1294073
Osenga, E. C. (2018b). iRON Soil Moisture Calibrated 2013 to 2018 (Version V1). Data set.
Zenodo. Retreived from https://doi.org/10.5281/zenodo.1271667

© 2019 American Geophysical Union. All rights reserved.

Pagano, T. C. (2010). Soils, snow and streamflow. Nature Geoscience, 3(9), 591–592.
doi:10.1038/ngeo948
Parida, B. R., & Buermann, W. (2014). Increasing summer drying in North American
ecosystems in response to longer non-frozen periods. Geophysical Research Letters, 41,
5476–5483. doi:10.1002/2014GL060495
Pecl, G. T., Araújo, M. B., Bell, J. D., Blanchard, J., Bonebrake, T. C., Chen, I. C., …
Williams, S. E. (2017). Biodiversity redistribution under climate change: Impacts on
ecosystems and human well-being. Science, 355(6332). doi:10.1126/science.aai9214
Peng, J., Loew, A., Merlin, O., & Verhoest, N. E. C. (2017). A review of spatial downscaling
of satellite remotely sensed soil moisture. Reviews of Geophysics, 55(2), 341–366.
doi:10.1002/2016RG000543
Robinson, D. A., Campbell, C. S., Hopmans, J. W., Hornbuckle, B. K., Jones, S. B., Knight,
R., … Wendroth, O. (2008). Soil Moisture Measurement for Ecological and
Hydrological Watershed-Scale Observatories: A Review. Vadose Zone Journal, 7(1),
358. doi:10.2136/vzj2007.0143
Seneviratne, S. I., Corti, T., Davin, E. L., Hirschi, M., Jaeger, E. B., Lehner, I., … Teuling, A.
J. (2010). Investigating soil moisture–climate interactions in a changing climate: A
review. Earth-Science Reviews, 99(3–4), 125–161. doi:10.1016/j.earscirev.2010.02.004
Seneviratne, S. I., Nicholls, N., Easterling, D. R., Goodess, C. M., Kanae, S., Kossin, J., …
Zhang, X. (2012). Changes in climate extremes and their impacts on the natural physical
environment. In: Managing the Risks of Extreme Events and Disasters to Advance
Climate Change Adaptation [Field, C.B., V. Barros, T.F. Stocker, D. Qin, D.J. Dokken,
K.L. Ebi, M.D. Mastrandrea, K.J. Mach, G.-K. Plattner, S.K. Allen, M. Tignor, and P.M.
Midgley (eds.)]. A Special Report of Working Groups I and II of the Intergovernmental
Panel on Climate Change (IPCC). Cambridge University Press, Cambridge, UK, and
New York, NY, USA, pp. 109-230
Service, N. R. C. (n.d.). Snow Telemetry (SNOTEL) and Snow Course Data and Products.
Retreived from www.wcc.nrcs.usda.gov/snow
Stohlgren, T. J., Falkner, M. B., & Schell, L. D. (1995). A Modified-Whittaker nested
vegetation sampling method. Vegetatio, 117(2), 112–121.
Udall, B., & Overpeck, J. (2017). The twenty-first century Colorado River hot drought and
implications for the future. Water Resources Research, 53(3), 2404–2418.
doi:10.1002/2016WR019638
Vano, J. A., Udall, B., Cayan, D. R., Overpeck, J. T., Brekke, L. D., Das, T., … Lettenmaier,
D. P. (2013). Understanding Uncertainties in Future Colorado River Streamflow.
Bulletin of the American Meteorological Society, 95(1), 59–78. doi:10.1175/BAMS-D-
12-00228.1
Whener, M. F., Arnold, J. R., Knutson, T., Kunkel, K. E., & LeGrande, A. N. (2017).
Droughts, floods, and wildfires. In Climate Science Special Report: Fourth National
Climate Assessment, Volume I. Washington, D.C. doi:10.7930/J0J8BNN
Williams, A.P., Allen, C. D., Macalady, A. K., Griffin, D., Woodhouse, C. A., Meko, D. M.,
… McDowell, N. G. (2012). Temperature as a potent driver of regional forest drought
stress and tree mortality. Nature Climate Change, 3(3), 292–297.
doi:10.1038/nclimate1693

© 2019 American Geophysical Union. All rights reserved.

Worrall, J. J., Marchetti, S. B., Egeland, L., Mask, R. A., Eager, T., & Howell, B. (2010).
Effects and etiology of sudden aspen decline in southwestern Colorado, USA. Forest
Ecology and Management, 260(5), 638–648. doi:10.1016/j.foreco.2010.05.020

© 2019 American Geophysical Union. All rights reserved.

Table 1. The above table describes instrumentation and general metadata for each of the
iRON stations. A “Standard Suite” consists of: 6-watt solar panel; 2 or 3 m grounded metal
tripod; Onset tipping bucket rain gauge; Onset relative humidity temperature probe in
radiation housing; Onset RX3000 Logger Box; Onset 12 bit soil temperature sensor (at a 20
cm soil depth); Decagon EC-5 dielectric soil moisture probe (5 cm soil depth); Decagon 10-
HS dielectric soil moisture probes (20 and 50 cm soil depths). Note: The Castle Creek station
is planned for installation in spring 2019.

Station Name Site Dominant Modified

(ID Number) Install Elevation Vegetation Whitaker Soil Characteristics Instrumentation
clay loam, 4.8% OM,
Glenwood Gambel
2015 1,890m 2017 3.0ppm NO3-N, 5.4ppm Standard suite
Springs (5) oak
P, pH 7.5
sandy loam, 33.6% OM, Standard Suite +
Glassier disturbed
2014 1,970m N/A 9.0ppm NO3-N, 100ppm soil moisture at
Ranch (3) wetland
P, pH 6.3 100cm
Standard suite +
pinyon, clay, 2.1% OM, 0.7ppm
Spring 2016, secondary set of
2016 2,160m juniper, NO3-N, 2.8ppm P, pH
Valley (8) 2017 soil moisture, soil
sage 6.5
temperature
Standard suite +
disturbed, clay, 6.8% OM 5.4ppm
Brush secondary set of
2014 2,370m brome 2017 NO3-N, 4.0ppm P, pH
Creek (4) soil moisture, soil
hayfield 6.5
temperature
Northstar sandy loam, 16.6% OM,
aspen
Aspen 2015 2,450m 2017 5.0ppm NO3-N, Standard Suite
stand
Grove (6) 20.5ppm P, pH 7.0
(20cm depth) sandy
loam, 10.8% OM,
3.1ppm NO3-N, Judd snow depth
mixed
Northstar 11.2ppm P, pH 7.4 sensor, soil
grass
Transition 2015 2,450m 2017 moisture (5cm,
transition
Zone (7) (50cm depth) 20cm, 50cm), soil
zone
sandy loam, 0.9% OM, temperature (20cm)
2.7ppm NO3-N, 3.8ppm
P, pH 7.6
loam, 10.6% OM,
Sky aspen
2012 2,550m 2016 60.5ppm NO3-N, 14ppm Standard Suite
Mountain (1) stand
P, pH 5.7
mixed loam, 0.9% organic
Smuggler
2013 2,759m conifer, 2017 matter, 0.5ppm NO3-N, Standard Suite
Mountain (2)
shrubs 30ppm P, pH4.7
Planned: Standard
Suite + snow depth,
Castle Subalpine
2019 3,290m N/A N/A radiation, wind
Creek (10) fir
speed/direction,
Reconyx Phenocam
Standard Suite with
Datagarrison
sandy loam, 7.0% OM,
Independence alpine satellite logger box,
2016 3,680m 2016 16.2ppm NO3-N,
Pass (9) willow Judd snow depth
18.4ppm P, pH 4.8
sensor, Reconyx PC
900 Phenocam

© 2019 American Geophysical Union. All rights reserved.

Figure 1 provides a map of the Roaring Fork Watershed, located in the Southern Rocky
Mountains of central Colorado. The 3,760 square kilometer Roaring Fork Watershed is part
of the larger Colorado River Basin, and Roaring Fork River is an important tributary of the
Colorado River. Figure credit: (Katzenberger & Masone, 2009, publisher permission has been
granted).

© 2019 American Geophysical Union. All rights reserved.

Figure 2. A map of the Roaring Fork Watershed, with iRON stations shown as red stars,
SNOTEL stations as blue circles, and snow course sites as yellow triangles. From left to
right, the iRON stations are: Glenwood Springs (5), Spring Valley (8), Glassier Ranch (3),
Brush Creek (4), Sky Mtn (1), Smuggler Mtn (2), Northstar Aspen Grove (6), Northstar
Transition Zone (7), Castle Creek (10) (planned for 2019 installation), and Independence Pass
(9).

© 2019 American Geophysical Union. All rights reserved.

Figure 3. An overview of the existing data record for iRON soil moisture, rain, and air
temperature sensors gathered since its establishment, ordered by elevational gradient, from
highest at the top to lowest at bottom. On the horizontal axis, each year is labeled on January
1. Because rain is measured by tipping-bucket gage, only growing season rain measurements
(May-Oct) are included in this graphic. The Northstar Transition Zone station is located at the
same elevation as Northstar Aspen Grove and is omitted for simplicity of presentation.

© 2019 American Geophysical Union. All rights reserved.

Figure 4. (a) Soil moisture (at 5, 20, and 50 cm) is shown on the y-axis, while date and time
are shown on the x-axis. (b) Total daily rain in cm and average daily soil moisture (at 5, 20,
and 50 cm) are shown on the y-axis, while dates from February 1, 2013 to Oct 31, 2013 are
shown on the x-axis.

© 2019 American Geophysical Union. All rights reserved.

Figure 5. Elevation (in meters) is shown on the x-axis for each iRON station. Station
elevation is plotted against “spring saturation” for the years 2017 and 2018, as defined by the
date of peak soil moisture in spring prior to moisture decline. The winter preceding spring of
2017 was a year with near average snowfall in the Roaring Fork Valley, while the winter of
2017-2018 has below average snowpack throughout the basin. Installation of a station within
the 3,200m range is planned for future network additions.

© 2019 American Geophysical Union. All rights reserved.