Sign Up for an All-Access Pass

News

Long range weather forecasts stink – a look back at the winter snow forecast for 2015-2016

YOU ARE BLOCKING ADS
Would you mind adding OpenSnow.com to the safe list?
Or, support OpenSnow directly for just $19/yr and enjoy no ads, 10-day forecasts, and unlimited custom alerts.
Advertisement

The snow forecast for the 2015-2016 winter was supposed to be somewhat simple. Nearly all forecasters agreed that we would see one of the strongest El Nino events of the last 50 years, and this El Nino should have decently predictable effects on the winter storm track over North America.

We did in fact experience one of the strongest El Nino events, but unfortunately,nearly every forecaster was wrong about the winter storm track.

Most forecasters did not correctly predict the overall weather pattern, which was supposed to provide a lot of rain and snow for the southwestern US and drier-than-normal conditions for the northwestern US.

Of course, there are exceptions to the ‘everyone got it wrong’ statement. For example, I think BA did a nice job being cautiously optimistic with his Tahoe forecast last season.

Regional predictions aside, if we concentrate on the general weather pattern, we’ll see that most forecasters did not make an accurate six-month prediction.

 

What actually happened

The maps below shows the actual snowfall compared to average.

  • Orange is below average snowfall
  • White is about average snowfall
  • Blue is above average snowfall

The maps present identical data in different forms. The dots represent ski areas and the colored states represent the average of all ski areas in each state.

The main take away from these maps is that much of the west saw close to or above average snowfall, the southwest experienced near to below average snowfall, and the northeast had a tough winter with well below average snow.

The data we used to make the maps is sourced from backcountry weather stations called SNOTEL as well as local volunteer reports incorporated into the Global Historical Climatology Network (GHCN). We like using this data because it does not contain any possibility of ski area marketing bias and because we can automate its collection.

 

Mid-winter switcheroo in the west

In the west, most forecasts pointed to above-average snowfall in the southwest.

The season did start strong in the southwest, and as of January 15, 2016, the snowpack was well above average (left side of the graphic below).

However, the second part of the season became much drier, and by April 1, 2016, the snowpack was well below average (right side of the graphic below).


I don’t know exactly why this switcheroo occurred, and no forecast that I saw accurately predicted it.

 

Holding the forecasters accountable

Below I show seven forecasts made during the early fall of 2015 for snowfall during the 2015-2016 winter.

All of these forecasts predicted generally the same weather pattern, with high-than-average snow in the southwest and well below average snow in the northwest. 

All of the forecasts, including ours, were wrong about this pattern. In reality, the southwest saw average to below average snow while the northwest did well, ending the season with near average or above average snow.

 

OpenSnow

We got it wrong (link to original forecast). The northwest was snowier than we thought, and the opposite was true in the southwest.

 

WeatherBell

Weatherbell produces forecasts for many industries and also is a source for weather data, providing forecast maps from nearly every weather model for $25 per month. We read their forecasts (and learn a lot from them) and also pay them for weather data.

Unfortunately, they got it wrong (link to original forecast). The northwest did much better than they predicted.

 

Accuweather

They got it wrong (link to original forecast). Again, the northwest was NOT drier than average, and the southwest was NOT much wetter than average.

 

NOAA

NOAA is the parent organization of the National Weather Service.

They got it wrong (link to original forecast) for the same reason as most of the forecasts above – the northwest was not significantly below average, and the southwest was not significantly above average.

 

OnTheSnow

Before launching OpenSnow in 2011, I used to write occasional articles for OnTheSnow. Now we are in direct competition, but I do appreciate their early support. For the last three years, neteorologist Chris Tomer has been writing forecasts for them. Chris is a TV meteorologist based in Denver. He’s a good forecaster and a globally-experienced mountaineer.

But, like the others, he got it wrong (link to original forecast) as the northwest had a average to above average season while the southwest suffered during the second half of the season.

 

Farmers Almanac

Looks like their secret formula was wrong (link to original forecast). They actually did OK in the west, but the northeast was not bitterly cold.

 

Slate

Eric Holthaus and I spent a week together at a weather conference in Boulder, Colorado during the summer of 2002 (before the iPhone … so long ago!). Eric is a nice guy, but I would kindly suggest to him, and anyone else making a long-range forecast, to not proactively call it “highly accurate”. Eric wrote an article for Slate titled “A Very Early Yet Highly Accurate Guide to This Coming Winter”. It was published on August 18, 2015. 

About snowfall in the Pacific Northwest, Eric wrote, “One place that probably won’t benefit from this winter’s El Niño is the Northwest. It’ll be another low-snowpack year…”.

As you can see from the maps above, this was wrong, and the northwest had a fine snow year.

 

What Professor Powder says about seasonal forecasts

Jim Steenburgh is literally the professor of powder. He is a Professor of Atmospheric Science at the University of Utah and wrote the book about how and where to find the best snow, called “Secrets of the Greatest Snow on Earth”. If you are addicted to powder like I am, you should buy his book. It’s meant for a general audience and is NOT technical.

Here is what Jim said about long-range forecasts in a blog post during the spring of 2016:

It's time to face facts.

First, the skill and utility of seasonal outlooks, whether they show weighting of the odds or pick an outcome (e.g., below average, average, or above average), remains limited for many applications.

I suspect we will see improvement in coming years and decades, but where we are in seasonal prediction today is about where we were in weather forecasting in the 1960s and 70s. The models are crude and we're relying too much on past analogs and human intuition.

Ultimately, these outlooks are going to improve, but even then, we must guard against converting those probabilities in to deterministic outcomes (e.g., increased odds of above average precipitation equals above average precipitation).

Second, we need to be more cautious about pushing analogs based on past events too far. A few months ago, many people were using the past two super El Ninos as analogs for this coming winter. A sample size of two is simply too small.

Finally, if you are skiing in northern Utah, you should just stop looking at seasonal forecasts altogether. SEASONAL OUTLOOKS ARE UTTERLY AND TOTALLY VALUELESS FOR ANTICIPATING WHAT KIND OF SKI SEASON WE ARE GOING TO HAVE IN NORTHERN UTAH. Next October, don't waste your time on this stuff. Wax your skis and be happy.”

 

Closing thoughts on seasonal outlooks

Some long-range forecasts can have value, but most will turn out to be incorrect. One day, perhaps in 20-30 years, this won’t be the case. But for now, 6+ month forecasts are not something that meteorologists do very well.

It is fun to look at the seasonal forecasts and learn about the logic that is used to create them. But don’t take the forecasts too seriously, and save them so you can look back in 6 months to see if they were accurate:-)

 

All Access

Love powder? We do too, and that’s why we created the OpenSnow All-Access Pass.

Being an All-Access member allows you to see our 10-day forecasts, time-lapse webcams for tracking exactly when fresh snow has fallen, and the ability to receive custom alerts to know when powder days are approaching. All of this costs just $19 for one full year (365 days).

In other words, one year of getting the best information about powder will cost you about the same as you’ll spend on one lunch at most resorts.

I’d love to count you as an All-Access member, and you can find out more here.

 

Email Alerts

If you’d like to get alerted when we publish future articles like this one, and also get access to deals and contests, sign up for our email list. It’s free, and you can unsubscribe at any time. Sign up by clicking here —> http://opsw.co/2bao4gA

 

Keep dreaming of future powder days!

JOEL GRATZ

 

 

YOU ARE BLOCKING ADS
Would you mind adding OpenSnow.com to the safe list?
Or, support OpenSnow directly for just $19/yr and enjoy no ads, 10-day forecasts, and unlimited custom alerts.
Advertisement

Recommended Content

YOU ARE BLOCKING ADS
Would you mind adding OpenSnow.com to the safe list?
Or, support OpenSnow directly for just $19/yr and enjoy no ads, 10-day forecasts, and unlimited custom alerts.
Advertisement
Please disable ad block or upgrade to ad-free.