#24: NoRRTN: An inexpensive regional river temperature network for the Northern Rockies

[Dan Isaak]

How $50,000 = 1,000 years of river temperature measurements (& informs
us about 20,000 years)

Hi Everyone,
Time to put my money where my mouth is after that last blog post
regarding the need for more long-term, full year temperature
monitoring of unregulated rivers and streams. So this week I want to
introduce NoRRTN to demonstrate just how easy & inexpensive it is to
set up a temperature monitoring network with modern sensor technology
these days. NoRRTN stands for Northern Rockies River Temperature
Network & is a project we started a few years ago that involved
instrumenting most of the large rivers across the northern Rockies
with full year temperature sensors (graphics 1 and 2). Not that there
isn’t some temperature data from these rivers already, but it often
consists of only a few summers measured sporadically here and there or
monitoring occurs in association with dams, reservoirs, and urban
environments that alter stream thermal regimes to some degree and
therefore don’t constitute the best climate sentinel sites (see blog
#23). A lack of temperature data from larger rivers is a significant
problem since these systems often constitute important recreational
fisheries and migratory corridors for salmon and other species that
move long distances. Moreover, when it comes to climate impacts this
century, these rivers are often already the warmest within a region,
and are where the rubber hits the road in terms of having high profile
resources at risk and strong public scrutiny. We need to be preparing
for that scrutiny by getting good thermal inventory data now as a step
towards developing the capacity to accurately model river thermal
regimes associated with climate change scenarios so that it’s possible
to see ahead of the next bend in the river and manage accordingly.

The first step in that process is simply getting more and better
temperature data from rivers and streams. We’ve developed our
preferred technique for monitoring full year temperatures that uses
underwater epoxy to glue sensors to large boulders or bridge pilings
(graphics 3, 4, and 5; blog #3; YouTube video demonstration @
http://www.youtube.com/watch?v=vaYaycwfmXs&feature=youtu.be), but it
ultimately doesn’t matter how it gets done, as long as it gets done.
The next step is choosing where it gets done, and identifying the data
gap on larger rivers across the Northern Rockies was done simply by
studying the locations of existing temperature monitoring sites using
a GoogleMap tool we developed last year (blog #4). This tool provides
a unified spatial index to most of the sites we’re aware of in the
northwest U.S. where full year temperature monitoring is occurring &
is intended to provide a means of coordinating monitoring efforts
among the dozens of resource agencies that are doing so. All the
NoRRTN sites have been added to the GoogleMap tool, & if you go to the
website (http://www.fs.fed.us/rm/boise/AWAE/projects/
stream_temperature.shtml), you’ll notice they’re just the tip of the
iceberg because there are now almost 3,000 sites with full year
monitoring efforts underway. In some cases, these sites are part of
centralized monitoring programs like USFS PIBO/AREMP, USGS NWIS, or
BOR HydroMet, but in many cases, they are from individual biologists &
hydrologists on national forests or with fish and game agencies doing
it on their own and committed to monitoring streams in their backyard.
Most of the sites on the GoogleMap were set up in just the last few
years, so monitoring efforts are rapidly expanding (still a ways to go
though with hundreds of thousands of stream kilometers out there). So
far, we’ve been developing and beta-testing the GoogleMap tool with a
primary focus on streams in the northwest U.S. but are slowly
expanding its geographic scope (& are happy to do so) as people share
with us their monitoring site coordinates from other parts of the
country. We still have some revising of the sites shown currently on
the GoogleMap to make it as accurate as possible before this next
field season, but will have a comprehensive set of site revisions up
by May/June.

Regarding the logistics & costs associated with NoRRTN. Basically, two
technicians drove around and did the installations over the span of a
few months in the summer of 2010 and the fall of 2011. We tried to get
2-3 replicates in most of the region’s rivers that didn’t already have
temperature monitoring underway. When the road trips were done, we’d
established 250 new sites on 80 rivers across Idaho, Montana, and
Wyoming. Those 250 sensors cost $30,000 ($120/unit), then throw in
some salary, per diem, & vehicle costs on top of that, and you get to
$50,000. We’ll no doubt lose some of those sensors to floods and
vandalism in subsequent years, but data recorded at the remaining
sites will tell us a lot about thermal regimes in the region’s rivers
and how these regimes are influenced by climate.

Validation work for the underwater epoxy protocol based on a large
scale field test initiated in 2010 suggests first year retention rates
are 85% in channels with slopes < 3% that are typical of large rivers
(graphic 6; Horan 2012.pdf document attached to this email). Moreover,
once a sensor installation weathers that first year’s snowmelt flood,
subsequent retention rates should be even higher. Each sensor has a
service life of 5 years (battery life and memory capacity), so
assuming 200 of our sensors stay where we put them that gets us to
1,000 years of direct temperature measurements. Those data can be
leveraged using modeling techniques like those described in previous
blogs to develop link functions with air temperature and stream
discharge records from nearby climate stations to reconstruct historic
thermal regimes (#7, #14, #23) or make projections about the future
(#23). Since 50 year historical air temperature and discharge
monitoring records are common in many places, it’s often easy to go
back that far with historical river temperature reconstructions, and
since many climate models project similar warming trajectories for the
first half of this century, it’s not too hard to project that far into
the future. Using previously published techniques, therefore, with
those 5 years of empirical temperature measurements could conceivably
provide inference about 100 years of thermal conditions at a river
site. 200 sensors x 100 years gets us to 20,000 years, so not too bad
for that initial investment of $50k.

Models will be important to use with monitoring data because they are
the only means we have of looking around that next river bend in time
to navigate the right line in the present. However, models are also
imperfect representations of the real world & so in some places at
least, it’s going to be important to commit to monitoring indefinitely
for the foreseeable future. For example, most of the best evidence we
have regarding how stream discharge regimes are changing in response
to climate forcing comes from 50 rather than 5 years of monitoring
(blogs #17, #18, #21). Moreover, monitoring & watching the same place
the same way for a long time helps catch the unexpected events that
may be especially illuminating with regards to revealing important
mechanisms and may also develop a deeper sense of appreciation in the
watcher.

For me, traveling around the region to install temperature sensors in
rivers has instilled a sense of awe in these ancient and dynamic
systems and the beautiful landscapes they are parts of (graphics 7 –
24). Today’s rivers have been here doing their thing for 10,000’s –
100,000’s of years and will be doing their thing for that long into
the future, but at present are changing more rapidly than they ever
have. My hope is that we’ll be able to maintain the NoRRTN network
over the next few decades of my career to better document some of
these changes, and that similar temperature monitoring efforts might
be initiated in rivers and streams everywhere. The better we can
describe how things are changing, the easier it will be to discern
underlying processes and to manage more effectively through this
transitional century. None of us will be here to know exactly what
rivers and streams are like a century from now when the climate change
story can be told more precisely in retrospect, but our data can help
future stewards of aquatic resources understand where we started and
the path that was traveled.

Until next time, best regards,

Dan