Surfing Safely during COVID

Many coastal communities sheltering in place have stepped in and closed beach access to all comers, surfers included. There’s been much frustration from surfers who’ve built their life around the beach and from those who feel there is a version of surfing right now that isn’t more risky that jogging or riding a bike. Any analysis of COVID at this stage is partial, we still don’t really know if, for example, aerosol transmission is possible or how or if the virus might survive in sea water, but some understanding of the data helps us understand and lobby for a responsible surfing that seems as low risk as is possible.

Socially distant surfing

The strongest argument right now is for socially distant surfing. The case here is simple: Surfing that respects social distancing guidelines seems likely to reduce the risk of infection to near zero. On many long, open beach breaks it’s entirely possible for many of us to get in the water and never come close to 6ft from each other, in fact it’s likely easier in many locations to do this than it is to jog or cycle alone on the sidewalk or road. If we’re suitably risk averse and don’t otherwise ask of our first responders by pushing our limits it’s hard to see a long term argument against a resumption of this kind of surfing.

But what about Australia?

Australia is one of a very few poster children for appearing to get to grips with its outbreak and many surfers will have noticed that for the most part beaches there remained open and in some cases distinctly un-distanced. Doesn’t this make the case that surfing, even clustered at popular locations, is low risk?

Simply put, no. This blog post explains really well how the risk of being in a group is relative to both the size of that group AND, crucially, the overall infection rate. For a simple demonstration here are some really rough estimates for the relative risks of rubbing shoulders with a crowd in a Californian versus New South Wales lineup using the same math:

You’ll notice that the rate the risk increases is dramatically different. 40 people in the water at Snapper might be a 15% risk of someone having COVID, the same 40 in California could be an almost certain risk. Crucially though this assumes that the risk is the same as for all of California and this is where things get complicated.

Why? Because cases cluster and we simply don’t know how many people have the disease for every one person who turns up at hospital needing a test. Dig into those California numbers and they look very different at the state, county and city levels:

The point here isn’t to explain your precise risk region by region, it’s to explain the difficulties in assessing it accurately given how variable the spread is and how sensitive a basic risk model is to that data. We’re not suggesting you start calculating your curve and trying to triangulate your risk, we’re just saying that curve can be really steep even in areas where the overall case rate doesn’t feel too scary.

So we’re back where we started —with the information we have now safe surfing right now is socially distanced surfing. Make yourself the zero risk on the left hand side of the graph, because unless you know how steep it is you have no way of knowing what your actual risk is and it could be a lot higher than you realize.

Assumptions

My intention here ISN’T to try to put concrete numbers to the risks of surfing in a crowd. It’s simply to show it’s unlikely to be risk-less, even when the number of infections in your area doesn’t seem alarming and the risk can vary considerably from region to region in ways that are hard to calculate easily.

• Total confirmed cases are taken from google summary on 28th April.
• Population taken from google summary.
• Total confirmed cases may include recovered people — I’ve not allowed for this and presume the stats are like for like. Overall risk might vary but comparative shouldn’t.
• I’ve assumed undetected cases are 10x test identified cases. This is one of the least understood and most impactful numbers and may be much higher than this, which would radically steepen curves. The latest antibody testing could suggest this is actually 5–8x higher than my estimate.
• It’s reasonable to assume some positive effects of selection bias here — that people who are sick are less likely to be surfing. However the data that shows that a large part of the infected population is able to transmit the disease while asymptomatic should mean this analysis is still meaningful.
• ‘Worst Case’ in the chart title means the % risk of infection here assumes that if you’re in the water with someone infected you will get infected. This clearly isn’t likely true, but I can’t imagine anyone would knowingly take that part of the risk onboard (eg. if you knew there was someone infected in a room you wouldn’t enter — unless you were an utterly selfless healthcare professional wearing PPE).
• Of course surfing, even in a crowd, doesn’t have to mean sitting elbow to elbow — it’s hard to imagine that it’s as or more risky than sitting enclosed in the same restaurant, at the same time it’s a challenge to control your distance in such a dynamic environment . The point is — we don’t know — but we have to presume the ‘6ft or more’ guidelines should be observed.