Everest 2017: Free vs. Paid Weather Forecasts

We all know that weather can make or break an Everest summit, or the entire season. With so many climbers this year, a period of bad weather could result in nightmarish crowds. Similarly, a long period of good weather will allow the crowds to spread out resulting in a positive experience for most.

As I recently cited, the normal range of suitable summit days ranges from 6 to 16 (average of 11 since 2001). That is where the winds are low (under 30 mph) and temperatures reasonable (higher than -20F). But if we have a repeat of only a few summit days similar to 2012 with only 4, the climbers are in for a very difficult season.

Expert Forecasts

The best guides will buy weather forecasts from long time, proven professionals even though anyone can surf the Web for a free forecast. So why pay? I asked long time Everest weather expert Michael Fagin of Everest Weather.

Michael provides forecasts to some of the major expedition groups on Everest and in the past has worked with Ed Viesturs on Annapurna and Jimmy Chin on Meru.

His business is diverse with about 30% coming from climbers.  He is also an expert witness in court and does a snow forecast for his local school district. Michael works with another meteorologist doing an agriculture forecast for Napa Valley in California and, finally, forecasts for engineering firms looking for rain and stormwater tests.

He works out of his home in Redmond Washington, USA. Michael starts his day early to review six forecast models and then interprets the weather pattern before applying 15 years of experience to filter out any bias. He told his hometown newspaper, The News Tribune about his methods:

I get up at 5 a.m. and get the coffee and the forecast models going. I’ll look at what members send me about the prior day’s weather and look at how much snow they got in the last 24 hours, noticing any winds up high.

I’ll look at the feedback, I’ll look at the different blogs. There are some weather stations to look at. Then I read the forecast models and spend about three hours doing that, then prepare the forecast. I have it ready by 9 a.m.

In the afternoon, I’ll take one more look at it and then send it out about 7 p.m. my time, which is 7 a.m. their time.

Now that we see what goes on behind the scenes, let take a deep dive into free vs. paid according to Michael:

There are many free computer generated weather forecasts available on the web for Mt. Everest and for most cities. For Mt. Everest I have been told that these sites are popular: Mountain Weather and Meteoexploration .  These are just two sites of many, and I’m not making any evaluation of them.

Before I get into some of the details of the computer generated forecast,  it is instructive to take a brief trip down memory lane. In the early 1970’s, two scientists from the US National Weather Service (NOAA) developed some complex equations that would predict key weather parameters like wind, temperature, precipitation, etc. This tool was named Model Output Statistics (MOS) and they now provide forecasts for a wide range of cities in the US and Canada that offer a seven day forecast.   Very simply put, this takes current weather forecasts and weather patterns and statistically looks at similar patterns for the same time of the year and provides a MOS forecast.

Fast forward to today and we have many computer generated weather products at our fingertips. NOAA has developed more sophisticated forecast products with a graphical interface where you can click a city or latitude and longitude.   These forecasts gather data from the Numerical Weather Prediction (NWP) models of the US based GFS (Global Forecast System), along with some of the newer models. Then, some adjustments are made by local meteorologists to develop the final forecast product.

Some countries are running sophisticated NWP models on Supercomputers, which can process 2.8 quadrillion equations per second.  These computers get data from land based weather stations, radar, satellite imagery, weather balloons and more. The amount of data that can now be processed is very impressive.

This progress in technology is being used in many industries with the wind energy industry  on the cutting edge. Many of these wind forecast models utilize “Neural Networks” which is beyond the scope of this article to explain in detail. But here it is in a nutshell.  The Neural Network’s goal is to have zero errors in the forecast. After every forecast is made, there is an automated evaluation to determine which weather parameters are the most important. For you weather nerds, check here for more information.

Also, IBM is using their expertise to provide more accurate forecasts with their recent purchase of The Weather Underground. Offering coverage on a worldwide basis, just plug in any city or area.

Back to Mt. Everest! Which begs the question: with the free weather sites available why should one pay for a forecast?

There are several reasons. First, on some of the Everest free forecasts sites they use only one forecast model. I know several sites that use just the US based GFS (Global Forecast System). There are times that this model can be inaccurate compared to some to the other models. I look at all of the following forecast models: US GFS Model, European Centre Medium -Range Forecast Model, Taiwan Central Weather Bureau GFS Model, UK Met Office Unified Model, US Navy No Gaps Model, and the Indian Meteorological Department WRF model. Then I assign a weighted average to help derive a forecast and give the  highest weight to the forecast model that shows the best degree of accuracy for recent model runs.

Another drawback of many of these free sites is you are getting the forecast output from just one data point (latitude- longitude). For instance, if you plug the name Everest into the web site, you might  get wind speeds forecasted of 20 mph (32 km. /hr.). However, let us say that 200 miles (320 km) to the west, the approximate distance  to Annapurna.  the wind speed forecast is 50 mph (80 km/hr.) Most  of the free sites do not indicate that difference.

Thus, if the  forecast is off by 200 miles, (the forecast of winds over Annapurna has now shifted over Everest on real time basis)  you will not get reliable information for the actual location you need. And you may be in for an unwelcome surprise. I base my forecast on multiple data points, and in this situation would communicate the chance of the winds going up to 50 mph in both text weather discussions and graphical presentation.

Another drawback of many free forecasts there is no probability assigned to the forecast. Sometimes I look at 7 models and get 7 different solutions. Thus I would have low confidence in the forecast and I would communicate that to my members.

Another key drawback of many of these free forecast sites is that they do not indicate if a major cyclone might form in the Bay of Bengal and potentially track towards Everest. This is definitely information that you will want to have! Many times I can see the formation of a cyclone 7+ days in advance and although the exact track is uncertain, I can alert my members.

Are all of these free sites for your mountaineering trips providing poor forecasts? No!! I wanted to point out some of the shortcomings. If you have a website that is providing you consistently accurate forecast then that is a great tool to use. If not, or if you need more of the features I’ve mentioned, there are meteorological firms like mine that provide custom forecast at a reasonable fee.

Thanks Michael. As we all now predicting the weather is always challenging and even more so on Everest. Wishing everyone a positive season this year throughout Nepal.

Climb On!
Alan
Memories are Everything