Back from AGU

Back from the AGU. Some great talks, met some very interesting people and had a couple of good suggestions. No one hated my poster even though the printers managed to absolutely destroy the photos with pixellation! :(

The poster is complete as a first draft...

Whew

Thanks to a useful tutorial found online, and the patient help of a GIS specialist, I was able (4th time lucky) to finally get HEC-GEORAS up and running and have started to process my survey files through that for HEC-RAS analysis. It has been slow, and at times frustrating, getting this going but now it promises to be much faster.

On the other hand I don't have any AGU poster mockups yet - maybe by Friday?

Frequency Analysis

I'm doing frequency analysis right now. My eyelid will not stop twitching. Are the two things connected?

Next week I hope to have some AGU poster mockups to post up.

The Inconvenience of Gauges

The only information we have on streamflow in many streams is that recorded at gauges, and yet, to make the measurement easier, gauges tend to be located in unnatural places - bridges, weirs, along riprapped banks etc.

While this ensures a stable cross-section, it's not so great for figuring out where the bankfull stage is at the gauge. So we have to go up- or down-stream to the nearest natural bank and hope that it is close enough that values of slope, Manning's n, etc. will remain relatively similar to those measured at the gauge so that we can transfer the hydraulic geometry.

High Energy

Setting up stream gauging at the Gold River site has been challenging. The stream is very high-energy and not a good site for a traditional post with gauge and PVC tube stilling well. the budget does not allow for a creekside well gauge as WSC typically uses. So, we had to get creative.

I brought my Bosch Annihilator powerdrill along and after some searching, found a relatively vertical rock face to bolt a 1.8m steel ruler to to serve as a staff gauge. Whether it will last through the winter storms in questionable, I am looking forwards to seeing if it can withstand high water loaded with moving boulders and LWD.

Next up was the installation of the pressure transducer and again it was bolting time. We took a small boulder, placed a bolt and zap-strapped the transducer to the bolt. Then I put the whole mess in the creek in a pool sheltered behind a bedrock section of the channel and built a little hutch of other small boulders around it and capped with a large flat rock. Hopefully it will be protected there. If not, we have a backup transducer ready to install.

Finally we did some salt dilution gauging. There are a lot of pools so this was challenging. First test, with 1 kg of salt over 35m of channel, got a conductivity peak of 90 μs/cm after 10 minutes or so and had not returned to background levels after 40 minutes. Second trial, with 2kg of salt over 15m of channel, got a peak of over 400 μs/cm and was back almost to background after 20 min.

On The Feeling of Being Transported Sediment

Scott Weston and I rented wetsuits and swam down Norrish Creek's lower canyon, between the two stream gauges. Although it was a fun trip the research goal was to identify sediment sources in the lower canyon and to determine if the lower canyon is subject to debris flows or if debris flows from the steep walls and tributary creeks deposit in the main channel. Based on our observations it seems that the latter option is the case, although debris flows from Sally Creek might reach the apex of the fan, below the canyon, on rare occasions.

Back to Work

The hydraulic geometry is all done and I am taking a break to do some hydrologic consulting work (see picture).

In good news, the beta HEC-RAS for ArcGIS 9.2 is out so maybe I can finally analyze my cross-sections if I can persuade it to crash less than frequently.

Hydraulic Geometry

I am currently working on hydraulic geometry relationships for my field sites. It's slow and steady work, but it is always gratifying to find locations where reality closely matches theory.

Further to the Previous Post

This is an aggrading stream. There is a large woody debris jam downstream and a massive rockslide about 1km upstream. Gravel of the size visible in the photo is accumulating in the bed and raising it. Therefore, I expect that Qb << Qe for this stream.

Five Possible Conditions

The effective discharge (Qe) and bankfull discharge (Qb) can equal each other and both have a return period approximately equal to the mean annual flood (MAF). This would be the normal condition for a hypothetical river in an equilibrium state.

There are four other conditions possible:

1) Qb frequent, Qe infrequent (Qe>Qb)
2) Qb infrequent, Qe frequent (Qb>Qe)
3) Qb and Qe both frequent (Qb and Qe < MAF)
4) Qb and Qe both infrequent (Qb and Qe > MAF)

Condition 1 represents streams with stable banks composed of relatively immobile sediment - an armoured channel with a low sediment supply, for instance.

Condition 2 represents what is probably a degraded channel, that has recently deepened itself, or that developed under a previous hydroclimatic regime, but is not armoured. Sediment moves through the channel frequently but the bank rarely overtops. It could also represent a fine gravel and sand bed channel with armoured banks (for instance, cut into glaciofluvial cobbles).

Condition 3 would likely be an actively aggrading glaciofluvial, meandering stream with bankfull discharge possible during any high-temperature melt event, or another similar condition

Condition 4 would be like condition 2 but with the additional constraint of condition 1 (low sediment supply). These would likely be the most stable channels.

An Interesting Factoid

Reid (2005) reported that for small alpine streams in BC, Manning's n varied by 5 orders of magnitude from the lowest flows (highest n) to the highest flows studied, as flow resistance increased at lower flows with more coarse sediment exposed.

What Would Being A Grain of Sediment Be Like?

Trying to get inside the state of a grain being transported here...

Bedrock Channel Streams

Here is a question I have been thinking about. In a bedrock channel stream, although the streambanks are not alluvial in nature, nonetheless, there is often a defined bank, either by vegetation or a morphometric or geometric break. Is there a characteristic discharge associated with these banks or is it just marking the last big flood to flow through and scour off the vegetation? I am not sure.



Five Mile Creek near Nelson is a creek I did not survey because it had an almost totally bedrock reach above the gauge.

In work news, I'm testing out ArcGIS 9.2 and seeing how it syncs with HEC-GEORAS

Foam and leaf in Kanaka Creek. Not all of my fieldwork was strictly technical. Sometimes I just took photos because I saw something interesting.

HEC-RAS 4 again

Well, we finished the streamflow part of the HEC-RAS 4 course and have started on the sediment transport portion. The good news: HEC-RAS 4 does the Wilcock bedload rating curve. The bad news - it is a crappy one-dimensional solution that produces results for a mixture that are the opposite of what Curran and Wilcock report in real life. Hmmmm.

After taking this course I think that I will be using it (HR4) to determine bankfull reach profiles and NOT using it for sediment transport calculations. Maybe for a backup incipient motion threshold calculation as a check, though. It might do that well enough...

HEC-RAS

I am struggling with HEC-RAS 4.0 beta. It does not seem intuitive to me. I downloaded it and played with it and realized there are thousands of ways to make it crash or get bad results.

Fortunately there is a course in Portland next Monday through Friday on HEC-RAS put on by River Restoration NW. You can bet I'll be there!

Sediment Analysis

This is an example of the raw images I'm using for grain size analysis in this project. This method allows for more rapid and more spatially distributed characterization of grain size. It was hard to do a proper Wolman pebble count on a sandbar measuring 2m on a side. Details of the method I am using are presented in Graham, Reid & Rice (2005).

LWD

Fred Touche measuring LWD (large woody debris) in Deer Creek near Castlegar in July 2006.

Right now I'm doing the last bits of data entry - 2007 LWD surveys from my handwritten notes into Excel. No one seems to have quantitatively linked natural LWD density to effective discharge before; Rob Millar demonstrated a change in the bankfull and effective dischages after LWD was removed.

First Post

This is my first post. I am starting this blog to have a quick website where contributors to my project can drop in from time to time and check on my progress, and perhaps see what issues I am thinking about as my research progresses.

Right now I have a project proposal (I'll post it here later as a PDF), I have gathered most of my data, processed most of my raw data and am getting ready, after two years, to start data analysis. I'm taking a course in HEC-RAS in Oregon next month which will be very helpful in the streamflow data analysis. I'm pretty excited about that.

Here is a pretty stream picture to look at. What's going on in this channel? This is just upstream of a large jam in the Albert River. It seems that the channel is aggrading, so the bankfull discharge frequency should be low and the effective discharge should be higher. Sediment is moving into this reach but not out. Once I crunch the numbers I'll know whether or not I'm right.