Caney Fork - 2. Bee Creek to Collins River

Caney Fork, Tennessee, US


2. Bee Creek to Collins River (Lower Upper Caney Fork)

Usual Difficulty I-III (for normal flows)
Length 27.1 Miles

River Description

Canoeing and Kayaking Guide to the Streams of Tennessee, B. Sehlinger and ...


Putting in at Scott's Gulf Road end we had to portage across the small island to get to deeper more open water and encountered a sharp right bend with a few rocks which might challenge beginners at this flow (1.4?).  After that it was mostly staightforwards class 2 then class 1 and then some 'forested' sections that collect logs and debris so some maneuvering is required even towards the very end. 

I would definitely run this again with less skilled paddlers, sit on tops, canoes but someone in the group with some class 2-3 skill should go first in the wooded areas to find the safe passage. 


Lots of big open pools, tall cliffs of pocketed limestone, and huge poplars in the river. 


At 1.4 there was ample current the whole way and during spring it could easily go down to 1.1 if you don't mind some rocky shoals.  In summer I believe it goes underground. 



StreamTeam Status: Not Verified
Last Updated: 2014-05-04 20:22:47


Rapid Descriptions

icon of message No rapids entered. If you know names, and locations of the rapids please contact and advise the StreamTeam member for this run.

User Comments

Users can submit comments.
December 31 2013 (1972 days ago)
Matt WellsDetails
Forum: BoaterTalk From: bankfull1 Dec. 31, 2012 I posted this on a thread below but it got buried,
and I thought after geeking out on this for a few years the things I've learned might help some
other boaters catch the Caney Fork in those times where adequate flows are uncertain. First of all,
while I didn't check the gauge Saturday, I am almost positive that Caney Fork ran at least 1.4, but
probably more like 1.5 or a little more. There are many factors in predicting flows on the Caney
Fork, and I wanted to comment on the different styles of predicting flows I've seen mentioned and
describe the approach that works best for me. First, it is awesome of cravincreeks to make a
googledoc and share that info with us, and we should contribute to that effort with visuals any
time we can.
I like the curves, and you should pay particular attention to three things: 1) Number of Data
Points: under the scale of science, these curves are surely in their moment of conception. More
input data is needed to increase accuracy. 2) R factor: look at the curve and you will see r
factor. The closer that number is to 1, the better the apparent correlation and the more useful it
could be to paddlers. The correlation to the Calfkiller is only a 0.6, and Daddys has an r factor
of 0.62. However the new Obed gauge at Adams Bridge has an r factor of 0.95! This is because both
the Caney Fork and the Obed start right next to each other in Crossville. 3) Using only one of
these, even the Obed with a 0.95 r factor is choosing to ignore other data. Use all three in suite
with rainfall totals/distribution to get the whole picture. Second, while DeRossett is a great rain
gauge to use, it is not always the best indicator. It definitely wasn't in this situation. The rain
gauge at DeRossett is west of the Caney Fork watershed. Most of the rain fell in the southern and
eastern portions of the watershed, as indicated by the rain gauge in Crossville and the rainfall
graphics on the NOAA site in the Meadow Lake Park area. So if you always just assume DeRossett is
the best indicator, you will short yourself on windows to boat the Caney Fork. Here is how I look
at predicting flows on Caney Fork, and this is the same methodology I apply anywhere else: 1) rain
gauges / graphics: First of all, the new estimated rainfall graphics available here are a game changer in predicting flows. While they are
theoretical estimates, I can safely say the error bars seem to be smaller than worth bringing up.
They are verified and weight adjusted based on actual rain gauges in the area, and this particular
link weaves multiple radar source data sets together. This stuff is high tech and has changed
everything for many boaters. Just bring the transparency down to the 30% range and zoom in to see
which watershed got how much rain. You will be able to see what parts of the watershed got rain,
and how uniform or not the rainfall was. This degree of resolution is necessary when you only have
a limited window to boat. With the long nights of winter coinciding with the plateau paddling
season, timing is everything and these graphics help you boat proactively instead of
re-actively(you can hit the road based on where the rain fell instead of waiting on slow to respond
river gauges). NOAA,AFWS, and a few other organizations put actual rain gauge info online. Most
gauges on the plateau are NOAA gauges and can be found here: There are 3 rain gauges I use to estimate rainfall
in the Caney Fork: a) DeRossett - just West of the watershed b) Crossville RAWS - just East of the
watershed c) Big Lick - SE of the watershed These three surround the Caney Fork, and give a fair
picture of how uniform the distribution of rainfall was. Caney Fork can run when 1 or 2 of these
gauges have little to no rain if the rain is localized enough. Very rarely are totals uniform
across the whole watershed. Sure it doesn't matter when everything blew up the day before and you
know it's gonna be running, but we know that. The difficulty lies in deciding whether to take a
chance when it is questionable whether there has been enough rain to get it high enough to paddle.
To just put all your cards in with one of these gauges every rain event is just going to increase
the potential for missing windows of opportunity. Assuming good baseflows, 0.5 to 1.0 inches of
rain should be enough. However if the Caney Fork just dropped out the day before, like in this
case, as little as 0.20 inches of well distributed rain will do the trick. 2) Stream gauges: Here
are the useful gauges in estimating flow in the Caney Fork: a) USGS Calfkiller (175 square mile
drainage)- most popular correlation, but least connectivity to the Caney Fork watershed. This gauge
is a decent correlation, but is most useful in determining whether the water table and karst of the
western plateau is charged and full or not. The Calfkiller dabbles in the same geologic conditions
as most of the gorge of the Caney Fork, whereas the Obed and Daddys encounter little to no karst.
The Caney Fork ups the ante on most other streams with the added complexity of this karst
topography (limestone, ie caves). If the water table is not charged, 2-3 times the amount of rain
that would normally get the Caney Fork going still might not be enough. Looking at historical
annual graphs, it is easy to see when the fill is met and surface base flows develop. On the
Calfkiller, you want to see over 1000 cfs for the Caney Fork to be running 1.4 or higher. The
response time of water draining the plateau and then meandering through the valley to Sparta where
this gauges is can be significant. A substantial portion of the watershed exists in the valley and
can reduce this gauges usefulness in predictions. Knowing where the rain fell and what is causing
the gauge to rise is important; see above on rain gauges and graphics. If this gauge has already
peaked, you can be sure the Caney Fork did as much as a day earlier, assuming no new rain. b) USGS
Daddys Creek (139 square mile drainage) - The daddys drainage meets the southeast side of the Caney
Fork drainage, and is useful in knowing what baseflows are doing on the top of the plateau. The
gauge is a ways further downstream in it's watershed than the put-in for the Caney Fork though, so
it often starts to rise after the Caney Fork (111 square miles at its put-in). This makes Daddys
more useful in estimating baseflow conditions for the Caney Fork than response flows to recent
rains. That considered, if Daddys is over 500 cfs, that is a good indicator there is adequate flow
in the Caney Fork. c) USGS Obed at Adams Bridge (90 square mile drainage) - this relatively new
gauge does wonders for knowing what is going on in the Upper Obed, but it does even more for
paddlers by helping tremendously in predicting flows in the Caney Fork. This gauge usually responds
to rain a little bit earlier than the Caney Fork put-in, so it is particularly useful in knowing
what is going on at the Caney Fork. In my experience, over 300 cfs on the Obed is a green light for
the Caney Fork. d) USGS Basses Creek at Lake Tansi (8 square mile drainage) - this gauge has a
narrow use, but is great at what it does. This gauge tells you in no uncertain terms if and when
feeder streams of the Caney Fork are responding to recent rain. This gauge will respond a bit
earlier than all the others listed since it is so small. Though its smallness makes it useful in
early indications, this same trait limits it's applicability; it is so small it doesn't tell you if
what happens there will happen elsewhere in the watershed. This is why you need to use rain
gauges/graphics as well as stream gauges to get the whole picture. Assuming the rainfall at Lake
Tansi is indicative of elsewhere in the watershed, all feeders of similar size will rise at similar
times to Basses, and then begin to move down the main stem to the Caney Fork put-in. Notice that in
listing these 4 stream gauges I started with the biggest first and then whittle down to the
smallest last. This is to bring to light the critical component of timing. After a blanket inch of
rain in the winter, Basses Creek will respond before any of the other gauges. The bubble of several
streams like Basses comes together and forms a stream like the Obed at Adams. Then the Obed at
Adams grows and cuts further into the terrain and eventually is the size of Daddys where its gauge
is. Finally, Daddys eventually builds enough watershed to be the size of the Calfkiller. Since the
water has to flow from small to big, the timing is offset accordingly. Basses responds first, then
Obed at Adams, then USGS Daddys, then Calfkiller. Lowland streams at the Calfkiller can cause it to
respond earlier than it's watershed would suggest though, so this should be taken into account. Not
only can you not just use one stream gauge, or one rain gauge, but you also can't use only one type
of data. You have to use rain and stream flow together to answer the important questions. Here's a
quick rundown of how I used all this to know that Bee Creek and Caney Fork would come up Saturday
morning: 1) I watched the NOAA rain graphics. While the whole western side of the watershed got
little to no rain, the area just southwest of crossville got over half an inch, and the graphics
showed a quarter inch or more from Meadow Lake Park south in the central to southern part of the
watershed. 2) Now that I knew where the rain likely fell, I moved on to looking at rain gauges for
verification. The DeRossett gauge showed next to nothing considering where the rain fell, but the
Crossville gauge showed a third of an inch with slightly less at Big Lick. 3) Next I waited for
Basses Creek to respond, since I knew it rained there. Basses soon started responding, so next I
watched for the Obed to start coming back up. Once this happened I felt pretty good about heading
over there. I knew where the rain fell, and that the little tribs were responding and the main
steam was starting to as well. 4) I woke up early and hit the road. Since the rain fell less in the
NW and more in the SE, and also since I like Bee Creek so much, I went to Bee instead of Caney
Fork. Bee had plenty of water to our delight. When we paddled up to the confluence with the Caney
Fork, there seemed to be more water coming from it than Bee, which is normal and indicates Caney
Fork responded to the rain as well . When I got home I noticed that Daddys finally had responded to
a lesser degree. I also checked the Calfkiller and saw that it barely digressed from it's slow drop
from the last rain. This squares with what I already knew about the lack of rain in the NW corner
of the watershed. Basses and the Obed essentially show how the little rain we got turned back the
hands of time a full day on the Caney Fork/Bee Creek system dropping out. So we caught a good run
in the watershed with the same levels from the day before. We didn't see anyone else out there. We
caught it by using all the tools available. And at the end of the day, our predictions and what
happened balanced. My friend even went back and ran Bee again today and it still had water. The
point is that there are way more factors in predicting flows than we'll every be able to
practically address. This post only addresses a few major ones. Other considerations like the
timing of the actual rainfall across a watershed are there for the considering. But we should
certainly make use of all the information out there. We should find out about it, logically figure
out what it tells us, and use it when it counts. The weather is so fickle and hard to predict with
great accuracy, but in these minimal rain situations we're actually trying to extrapolate upon what
is already a daunting task of predicting weather. We are trying to predict how a product of weather
will behave and when. It is certainly complicated, but just like weathermen, we have to use all
available resources and be open to new ones as they come along. Just as weather prediction has
gotten much more accurate and precise, so can our flow prediction abilities. NOAA and USGS are
doing all the work for us, all we have to do is pay attention and be thankful we have access to the
data. I totally geek out on this stuff, but there is a reason. I don't get to paddle like I used to
when I was just a college kid with an addiction. I have plenty of responsibilities and little time.
So when I have a 3 or 6 or 9 hour window to run the best possible run going, I use all the tools
available to hedge my bets towards success. Figuring all this stuff out has taken years of failure,
thought, and number crunching, but nowadays I put in little effort to figure it out, because I have
put in the time. You don't have to put in the time like I did, just take what I've figured out and
roll with it. No point re-inventing the wheel, but there is certainly no use in just riding around
on a donkey. Surf the net and find the gold on the information superhighway. Kirk CaneyFork.pdf -
Google Drive

Do more than just check gauges; join over 5,000 AW members today.

Or, consider donating