Overview of Indirect Measurements - Survey Requirements

Video Transcript
Download Video
Right-click and save to download

Detailed Description

This video provides an overview of the different indirect measurement methods used by the USGS, as well as a brief description of what you'll need to survey for each. 


Date Taken:

Length: 00:08:11

Location Taken: Henderson, NV, US

Video Credits


Additional camera work of floods by: Katie Earp, Tony Carson, and Ron Spaulding. Closed-captions by Ruby Hurtado.


Hi, this is Megan Poff and I’m the Field Office Chief at the USGS in Las Vegas, Nevada.  When we need to make a discharge measurement we usually go to the site with our acoustic or mechanical meter and make a direct measurement, right?  However, during floods you might not be able to make a direct measurement.  The bridge might be out.  There might be too much dangerous debris to put anything in the water.  The flood might have occurred at 2am and lasted for an hour.  These types of high water events are usually very important parts of a stage/discharge relationship though.  Just because we can’t make a direct measurement doesn’t mean we give up on defining the rating.  There are alternatives for making a direct measurement, and this video series focuses on indirect measurement methods.  I’m going to talk about the types of indirect measurement methods you might decide to use in the field and provide a brief description of the data you’ll need to gather for each one.  If you are going to survey one of the types of indirect measurements I discuss in this video, please refer to the specific videos on each method for more guidance as well as the various Techniques of Water Resources Investigations (TWRI) reports that have been written for each of the methods.

Let’s discuss the main types of indirect measurements that are presented in this video series.  In no particular order, they are: slope-area, slope-conveyance, culvert computation, road overflow, contracted opening, critical depth, and combined methods, where you might employ two or more of the aforementioned methods to come up with a total discharge for the peak flow event.  You will need to flag high-water marks for all of these methods.  If you don’t have high-water marks available, you could consider a different type of computation called a step-backwater computation.  The step-backwater technique is outside of the scope of this video series.

The most common indirect measurement method employed by the USGS is slope-area.  Generally, if you’ve heard people in your office talk about “running indirects” they are most likely referring to slope-area measurements.  Here is what you need for a slope-area measurement: a long, straight reach with at least half a foot of fall, no major contractions or expansions, and no major changes in channel geometry and slope.  For example, don’t use slope-area methods when you have a weir or road embankment in the middle of a reach.  The total length of the reach should be at least three times the width of the flood peak channel.  If your flood channel is about 500 ft wide, your total reach length should be about 1500 ft at a minimum. 

Let’s move on to a slope-conveyance computation.  The requirements for a slope-conveyance computation are similar to those for a slope-area, but the reach length doesn’t need to be as long.  If the flood channel is about 500 ft wide, you’ll only need about 500 ft of reach length for a slope-conveyance.  However, a word of warning: if you have the choice of doing a slope-area or a slope-conveyance, choose the slope-area because the solution will be considerably more robust.  A slope-conveyance only uses one cross section, and unless that cross section is representative of the entire reach, the estimate of flow that it provides could have considerable error.  A place where you might consider using a slope-conveyance computation would be a uniform concrete channel.  Another place where you might use a slope-conveyance would be a natural channel that is only uniform for a very short distance.  If at all possible though, try to find a longer reach so you can do a slope-area computation instead.

Let’s move on to the culvert computation.  You could use a culvert computation if you have all of your flow contained within a culvert or culverts.  Make sure that you had ponding on the upstream side of the culvert.  You’ll be able to tell this in the field by taking a step back and looking at the upstream high-water marks as a whole.  If they all appear to be about the same elevation, then you probably had ponding.  Your reach will extend upstream about twice the width (including wingwalls) of the culvert opening.  The culvert itself will be included in the middle of the reach, and on the downstream side you’ll want to make sure you have a distance of about twice the width of the culverts.  High-water marks need to be surveyed both upstream and downstream of the culvert as the approach and tailwater conditions are critical pieces of information to determine peak flow through a culvert.  If you notice fill in the culvert, see if you can figure out if it was there during the peak or not.  If you don’t know the status of the fill, then a culvert computation might not be the right indirect measurement choice.

Let’s talk about road-overflow computations now.  As the name implies, you could do a road-overflow measurement when you had flow going over a road.  You’re basically using the road as a broad crested weir to compute a discharge.  Just like with a culvert, you need a reach that will encompass the road overflow on the upstream and downstream sides.  Also like a culvert, you need to make sure you have ponding on the upstream side of the road.  One extra thing to remember for a road overflow - the road has to have a crest or a high point.  Typically the crest is on the centerline, but crests could also be located on the upstream or downstream sides.  If the road doesn’t have a crest, a road overflow computation won’t work. 

What about contracted opening computations?  A contracted opening computation can be done when you have a structure like a bridge that contracts the flow, causing it to accelerate.  The reach requirements are the same as a culvert where you’ll need an upstream distance of twice the width of the contraction and about the same length downstream.  However, you need to be careful that the flow remains in a subcritical flow regime through the contraction, and also that the flow didn’t cause significant scour in the contracting section.  If you are unsure about any of the above, it might be wise to consider a different computation. 

Let’s talk about critical depth computations now.  These computations are great in places where you have a waterfall or some other well-defined drop where the flow clearly passed through critical depth.  You could even do a critical depth computation off of a structure that resembled a sharp crested weir as long as there was no submergence.  Make sure you have a free getaway and no backwater.  The reach length will need to be about twice the width of the flood channel upstream and about the same distance downstream.

Last but not least, I want to briefly go over combined methods.  The most common type is when you have flow going through a culvert or bridge as well as over a road on top of the culvert or bridge.  The computation itself can get pretty complex, but the good part is that the reach length requirements are the same as for a culvert or contracted opening.

So – let’s say you’re looking at a site where you have a flow going through a culvert OR you could go half a mile upstream to a nice slope-area reach.  Assuming no major gains or losses between the sites, how do you choose which method to use?  By “no major gains or losses,” I mean that you shouldn’t have another significant inflow or a large floodplain that will attenuate significant amounts of water.  Most of the time, it will sort itself out for you easily – some flow will have bypassed the culvert, or you weren’t sure of the fill levels in the culvert beforehand, or maybe the slope-area reach doesn’t have that magical half a foot of fall.  In this example though, none of that happened and both sites look great.  So what is the answer?  Well…I hate to tell you this, but the answer is “it depends.”  A lot of times, culvert computations can be more accurate than a slope-area, which would lead me to want to choose the culvert computation over the slope-area.  However, consider the site conditions.  What if your slope-area reach is in a concrete channel and the culvert reach is in a natural channel?  It’ll be a lot easier to determine the n-values in a concrete channel than that natural channel, so in a case like that I’d pick the slope-area method.  It isn’t just about n-values though.  You also have to consider the quality of your high-water marks.  If the high-water marks in the concrete channel are plus or minus a quarter of a foot but the high-water marks in the culvert are plus or minus a tenth of a foot, I’d pick the culvert computation!

What about a case where you have something more complex – say, a multiple-component culvert plus road overflow, versus a slope-area?  Again, it depends.  The multiple component computation could be more accurate than the slope-area, but the computation itself will be significantly more time-consuming.  If everything else is equal, balance the need for more accurate data with the availability of personnel to work, check, review, and approve the computation. 

If you plan on trying any of these computations, see the individual videos on each type.  Remember, if you need help in the field, call your supervisor, surface-water specialist or indirect measurement specialist.