Risk Assessment for Rivers and Avalanches

Disclaimer: This discussion is based on my personal experience, collaborators’ feedback on The Packraft Handbook, and publications (see references at bottom). This discussion benefited from the feedback of Trip Kinney, Heather Thamm, Chris Erickson, Betsi Oliver, Huw Miles and others on the email list, but the opinions are mine.

Luc Mehl, Cascadas Micos, Mexico

Trip Kinney, one of The Packraft Handbook‘s peer reviewers, commented: “You need to define risk.” So, I sat down and was like, no problem, “Risk is …. risk… is… hmm.” Then I googled it.

I ended up in a deep dive learning about risk as discussed by the avalanche and natural disaster communities. The avalanche literature was particularly helpful and relatable; it feels the most similar to the river setting and is familiar given my training and personal experience in avalanche terrain (see: Mt. Logan traverse).

Part of the problem when discussing risk is that many terms are similar: hazard and danger, exposure and vulnerability, etc. Another Packraft Handbook reviewer said, “I don’t like the use of these terms because I can think of three different interpretations for each one.” I agree, but then I realized that was part of the problem. Adopting the definitions designated by risk professionals will make it easier to discuss risk before and during our outings.

Risk is the likelihood of harm due to exposure to a hazard. You can assess risk with these questions: 

  • What can happen?
  • How likely is it that it will happen?
  • If it does happen, what are the consequences?

The degree of risk depends on hazard, exposure, and vulnerability.

  • Hazard: Hazards are environmental dangers, the things we can’t control.
  • Exposure: Exposure refers to the people and possessions that the hazards can influence.
  • Vulnerability: Vulnerability is the likelihood that exposure to a hazard will have harmful consequences.

I appreciate this hazard/exposure/vulnerability framework; it helps me make sense of my previous decisions and intuition regarding risk in the outdoors. I pictured these factors as dials on a mixing board and asked Sarah Glaser to illustrate them for The Packraft Handbook. I got some sweet tattoos out of the deal (cover photo).

Probability and Consequence

The hazard/exposure/vulnerability framework works well for big-picture decision-making: where to go and what to bring. But when we are actually on the river, a simpler assessment is preferable. On-the-fly risk assessment boils down to probability and consequence:

  • Can I stay upright?
  • How bad will it be if I don’t?

Probability and consequence are treated differently by the avalanche and river communities. The probability and consequence of an avalanche are part of the hazard—things we can’t control. We can’t control the likelihood and size of an avalanche.

In contrast, the probability and consequence of swimming a rapid depends on the swimmer—a swimmer without a life vest is more likely to experience harmful consequences than a swimmer with proper safety equipment. In river settings, probability and consequence are part of exposure and vulnerability—things we can control, or at least influence. (With the notable exception of dynamic river hazards like floods and moving logs, which have probability and consequence like avalanches.)


Another way to describe the difference between avalanche and river risk assessment is to be explicit about the danger. Hazard and danger are synonymous in avalanche terrain. The slide is just as dangerous to anyone caught in it. An airbag could keep you on top of the debris, but the airbag changes your vulnerability, not the danger of the slide.

Hazard and danger are not the same on the water. The danger depends on the paddler and their equipment. As with the previous example, a swim is more dangerous for someone without a life vest.

Real vs. Perceived Risk

The real and perceived risks of an activity aren’t always the same. In packrafting, two of the most likely perceived risks identified by expert boaters and swiftwater rescue instructors are strainers (objects that let water pass through but not bodies) and foot entrapment (getting a foot stuck on the river bottom while standing in swift water). However, there have not been any known packraft fatalities due to these risks. Statistically, the most significant real risks involved with the twelve known packraft fatalities are cold water and paddling solo. Packraft safety courses should be emphasizing drysuits and paddling with capable partners.


We would all benefit from adopting standard definitions of risk assessment terms to the degree that it is possible. Risk in terms of hazard, exposure, and vulnerability provides an excellent framework for decision-making during trip-planning. On-the-fly assessment is easier if you focus on probability and consequence: How likely is something to go wrong, and how serious are the consequences.

Your willingness to expose yourself to risk depends on your risk tolerance. That sounds like a great discussion for a later post.

These two examples might be of interest to people who want to keep exploring the hazard/exposure/vulnerability framework:

Playing with the Dials: Avalanche Terrain

Luc Mehl, Mt. Logan

In avalanche terrain, the hazards are the snowpack, weather, terrain, etc. We can choose to expose ourselves to a slope or not; exposure is kind of an on/off switch that we control. Vulnerability is our susceptibility to being harmed. We have a lot of influence regarding vulnerability; we can reduce our vulnerability with training, practice, releasable bindings, and capable partners.

Hazard, exposure, and vulnerability are like dials that can be adjusted to match our risk tolerance. Take avalanche airbags as an example. Avalanche airbags can be inflated to make you more buoyant and less likely to get buried. Using an airbag reduces your vulnerability: the consequences should be less harmful if caught in a slide. If your risk tolerance remains constant after buying an airbag, your overall degree of risk is reduced. But a concern with airbags is that your tolerance might change. You might think, “I can ski this slope; the airbag will keep me safe.” In this case, turning down the vulnerability dial is countered by turning up the exposure dial, making you more likely to get caught in a slide (greater degree of risk).

Playing with the Dials: Rivers

Shasta Hood, Matakitaki River, New Zealand

River hazards are rocks, holes, waves, water temperature, etc. Exposure can be an on/off switch (paddle or portage) but may depend on your boat control skill (can you maneuver around the hazard). Vulnerability is our safety net: equipment (life vest, drysuit, safe rigging on the boat?), training and practice (able to perform a wet re-entry, deploy a throw rope?), and paddling with capable partners is a huge boost to the safety net.

We recognize that the vulnerability dial is cranked up during remote trips in Alaska. No one is available to help if something goes wrong, and a small accident (a swim in glacial water) could have harmful consequences (hypothermia). We choose to compensate for the increase in vulnerability by decreasing our exposure: committing to portage difficult rapids and reducing our risk by choosing rivers with fewer hazards, say, Class II water instead of Class III.



  1. thanks, Luc, your definitions, examples etc make this complex subject easy to understand!

  2. Just some FYI. You’ve likely searched through reams of this stuff.

    Federal agencies are continually assessing risk. Maybe some of their tools will help you?

    NCPTT | Risk Calculator App https://www.ncptt.nps.gov/blog/risk-spe-orma-and-gar-calculator/


    Risk Management | US Forest Service http://www.fs.usda.gov/managing-land/fire/safety

    You’ve likely looked at all the mountain rescue stuff too…just FYI


    On Fri, Jan 15, 2021 at 6:59 PM Things To Luc At wrote:

    > lucmehl posted: ” Disclaimer: I am not trained as a risk assessment > expert. This discussion is based on my personal experience, collaborators’ > feedback on The Packraft Handbook, and publications (see references at > bottom). This discussion benefited from the feedback of Tr” >

    1. Thank you Lisa! I didn’t find federal and natural disaster risk assessment to be as relevant to recreation because with recreation we are so clearly choosing to be exposed.

      But yes, it is helpful to see these, and I appreciate that you’ve compiled these references for folks that want to dig deeper.

  3. Awesome post! I’ve been a fan of your adventures for some time now and always appreciate a look behind the pretty pictures! I love participating in safety culture and feel empowered to work more and more of it into my guiding and instruction in the outdoors and around the shop.

    To your HEV concept, I’ve always practiced Severity x Probability x Exposure, same concept, but yelling “SPE!” to my fellow guides above a rapid is maybe easier than “HEV!” haha. If you haven’t looked at GARs, green amber red, there’s some good concepts to incorporate into your trip planning, especially with a team of diverse experiences.

    One big idea I haven’t seen really talked about much, since Alex Honold free soloed El Cap at least, is the breakdown of offensive/defensive optimism/pessimism. Sounds like you might be saving those ideas for another post though!

    Thanks for the great writing!

    1. Thanks for your contribution here Amir!

      There are certainly a lot of ways to address risk, different combinations of the same concepts. Your SPE is my “probability and consequence.”

      One thing I’m exploring is a pretty distinct difference between assessment during planning vs. assessment on the water (or scouting). I think we are unlikely to do a good job with HEV or SPE on-the-fly. In those settings, I think we want to make things as simple as possible: Can I stay upright? How bad will it be if I don’t.

      I’m looking forward to the summer paddling season to put some of this to practice!

  4. One thing I’ve done some thinking about with respect to risk and probabilities is that humans don’t deal very well with large numbers, even somewhat large numbers. For instance, if I asked you to estimate how much time you spend per year brushing your teeth without calculating, I suspect you might be surprised. 4 minutes per day equates to 24 hours a year – a full day (including sleeping) spent scrubbing your teeth.

    Similarly, if I asked you for the maximum acceptable risk of serious injury for walking up and down stairs in your house, that turns out to be substantially lower than you might guess. If I assume that I go up and down the stairs in my house 20 times per day, and I want to have only a 10% chance of being seriously injured by those stairs over a 20 year span, I come up with a maximum probability for a single trip up and down the stairs of 1 out of 1,400,000 (roughly). Another way of putting it – if I have a 1 in a million chance of serious injury from something I do 20 times per day, then over a 20 year span, I have a 13.6% chance of serious injury. Now I don’t feel at all like have any way of looking at an activity and estimating the probability of serious injury (outside of finding statistics) – is it a 1 in 10,000 chance vs. 1 in 100,000 vs. 1 in a million vs. 1 in ten million? What I do know is that when I think about this instinctively, my brain automatically ups the probability of injury, probably because my brain already knows that I’m going to underestimate how many times I will do something over my (hopeful) lifespan.

    1. This reminds me of some feedback I received about the necessity of separating risk assessment into the rational part of our brain and the emotional part. It is easier to use the rational part when planning and without a sense of urgency. As soon as there is urgency, we resort to an emotional response.

      Thanks for your comment.

Leave a Reply