Smarter snowmaking prediction on home page
13 posts
4 users
2k+ views
Scott - DCSki Editor
January 7, 2005
Member since 10/10/1999 🔗
1,252 posts
Hello,

I have just rewritten the algorithms that compute current snowmaking conditions on the home page of DCSki (in the upper righthand corner, titled "Right now:".)

For your default weather location (which you can change by visiting the Weather page), the new algorithm looks at the most recent current conditions reported by AccuWeather (usually updated hourly) and assesses temperature, humidity, and wind speed to determine whether snowmaking is possible, and if so, how optimal conditions are. It displays an appropriate message and icon.

Keep in mind that temperatures can still vary (for example, the closest weather reporting station to Snowshoe might not have the exact conditions as the top of Snowshoe mountain). But I believe this new algorithm will be pretty darned accurate most of the time. I referenced tables that snowmakers use to optimize snowmaking. So don't be surprised if the current temperature is below freezing, but the home page reports snowmaking is not currently possible -- if it's too humid, snow isn't going to come out of the guns.

The 5-day forecast (which you get once you click to the detailed Weather page) uses my traditional snowmaking prediction algorithms that are based on a longer-term forecast. They won't be as accurate as my model based on real-time temps, but should continue to give you a good general idea of what snowmaking looks like the next few days.

- Scott
KevR
January 7, 2005
Member since 01/27/2004 🔗
786 posts
That's cool Scott - I wonder, would you consider publishing your algorithsm -- OR alternatively providing a class libary ... I still want my fancy animate snow gun, moving graphs multiple resort display! I could just do it for fun and run it as little app on my box... Or you could and put it out there. Or something.

skier219
January 8, 2005
Member since 01/8/2005 🔗
1,318 posts
Yeah, I would be really interested in the algorithm too -- I'm a NASA engineer and would like to learn how this is computed.

thanks,
Craig
Scott - DCSki Editor
January 8, 2005
Member since 10/10/1999 🔗
1,252 posts
Here's a good resource I found:

http://makesnow0.tripod.com/wetbulb.html

It's a wet bulb conversion table that shows the relationship between humidity and temperature and how it affects snowmaking.

- Scott
ts
January 8, 2005
Member since 10/19/2004 🔗
6 posts
Hi Scott,

If you're looking for more data points for input into your algorithm, check out http://www.rap.ucar.edu/weather/surface/ . From this site, you have a couple of options for viewing surface METAR observations.

1. You can click on the default map and view the data for a particular region in station model format.

2. You can create a larger map from which you can zoom in for more station model observations (note link above the default map). If you place your cursor over a station model, you'll get coded METAR data.

3. If you want to see just the coded or decoded METAR data for a particular station (up to 60 hours worth of data), use the text observation link located on the RHS of the page. You'll need the station identifiers, but notice that there is a link to a text file that contains those identifiers.

-Todd
KevR
January 8, 2005
Member since 01/27/2004 🔗
786 posts
I could see an app that sat on your desktop, had the weather gun predictor for your, or any location you choose, had a snow dump news wire at the bottom, had adverts pushed to it from snow places or snow businesses, and scrolled the current topics in the dcski discussion group -- so you could keep up with gossip. Why I'd pay up to 0$ for this but hey, someone might pay, or maybe the adverts would make u some $. IF NOT, and jettisoned the money side, just freeness of of it would be cool by me.
skier219
January 9, 2005
Member since 01/8/2005 🔗
1,318 posts
So it turns out that it's all based on the wet bulb temperature, which you can compute from ambient temp and relative humidity. Wet bulb temp less than 20F is excellent (dry snow), 21-27 is OK (wet snow), and 28-> is no snowmaking. There's a wet bulb temp calculator web page (and a link to a spreadsheet) here:

http://www.the-snowman.com/wetbulb.html

I wrote a little UNIX shell script that pulls weather info off the National Weather Service XML feed and displays the info along with the snowmaking condition. I'll probably refine it over the next couple days and maybe wrap it in a GUI. It could be a little app that sits on the desktop and keeps tabs on favorite ski areas.

CRaig

Sample Output:

Current Snowshoe Weather
--------------------------------------------------------------
Report: Last Updated on Jan 8, 11:51 pm EST
Conditions: Overcast
Temperature: 37 F
Rel. Humidity: 79 PERCENT
Wet Bulb: 34.6533 F
Wind: Calm
Gusts: 0 MPH
Wind Chill: Not Applicable F
Snowmaking: NOT POSSIBLE

Current Wintergreen Weather
--------------------------------------------------------------
Report: Last Updated on Jan 9, 12:00 am EST
Conditions: Fair
Temperature: 34 F
Rel. Humidity: 87 PERCENT
Wet Bulb: 32.6206 F
Wind: Calm
Gusts: 0 MPH
Wind Chill: Not Applicable F
Snowmaking: NOT POSSIBLE

Current Sugarbush Weather
--------------------------------------------------------------
Report: Last Updated on Jan 8, 11:51 pm EST
Conditions: Fair
Temperature: 19 F
Rel. Humidity: 92 PERCENT
Wet Bulb: 18.3905 F
Wind: Calm
Gusts: 0 MPH
Wind Chill: Not Applicable F
Snowmaking: EXCELLENT (DRY SNOW)
KevR
January 9, 2005
Member since 01/27/2004 🔗
786 posts
sounds like u are onto something, will you publish?
skier219
January 11, 2005
Member since 01/8/2005 🔗
1,318 posts
Quote:

sounds like u are onto something, will you publish?




Yeah, I'll be glad to. You can also get a measure of the snow lifetime with the dry bulb temp (ambient temp). It could get very detailed, but it seems adequate to break it down into categories like good, fair, poor, etc....

I started making a little app for Mac OS X, since it's really easy to develop on that platform. Once the app is ready, maybe someone else can work up a Windows port. The basic shell script at the heart of it will run on any platform as-is. I'll post that tonight or tomorrow.

Craig
KevR
January 11, 2005
Member since 01/27/2004 🔗
786 posts
i was thinking of a java thing but no real plans
skier219
January 12, 2005
Member since 01/8/2005 🔗
1,318 posts
Quote:

sounds like u are onto something, will you publish?




Here is the code of the shell script:

Code:
  
#!/bin/zsh -f

FILE=/tmp/ft.xml

function wxgrab
{
key="<$1>"
rep1="s/<$1>//g"
rep2="s/<\/$1>//g"
grep $key $FILE | sed $rep1 | sed $rep2
}

function wet_bulb
{
TEMPK=$(((TEMP-32)/1.8))
echo $((((-5.806+0.672*TEMPK-0.006*TEMPK*TEMPK+(0.061+0.004*TEMPK+0.000099*TEMPK*TEMPK)*RH+(-0.000033-0.000005*TEMPK-0.0000001*TEMPK*TEMPK)*RH*RH)*1.8)+32))
}

function snow_life
{
echo $((0.027*0.027*TEMP*TEMP))
}

XMLLINT_INDENT=
export XMLLINT_INDENT
curl -s http://nws.noaa.gov/data/current_obs/$1.xml > /tmp/temp.xml
xmllint --output $FILE --format /tmp/temp.xml
rm /tmp/temp.xml

LOC=$(wxgrab location)
RH=$(wxgrab relative_humidity)
TEMP=$(wxgrab temp_f)
TIME=$(wxgrab observation_time)
WTHR=$(wxgrab weather)
WINDDIR=$(wxgrab wind_dir)
WINDSPD=$(wxgrab wind_mph)
GUST=$(wxgrab wind_gust_mph)
CHILL=$(wxgrab windchill_f)
TWB=$(wet_bulb)
#SL=$(snow_life)
SL=$TEMP

SNOW="EXCELLENT"
if [ $((TWB)) -gt 20.9 ]; then
SNOW="GOOD"
fi
if [ $((TWB)) -gt 27.9 ]; then
SNOW="NO"
fi

LIFE="EXCELLENT"
if [ $((SL)) -gt 24.0 ]; then
LIFE="GOOD"
fi
if [ $((SL)) -gt 32.0 ]; then
LIFE="FAIR"
fi
if [ $((SL)) -gt 37.0 ]; then
LIFE="POOR"
fi

printf "\n"
printf "%s\n" $LOC
printf "%s\n" $TIME
printf "Observation: %s\n" $WTHR
printf "Temp: %s F (Wind Chill: %s F)\n" $TEMP $CHILL
printf "Relative Humidity: %s %%\n" $RH
printf "Wind: %s %s MPH, gusting to %s MPH\n" $WINDDIR $WINDSPD $GUST
printf "Wet Bulb Temp: %g F\n" $TWB
printf "Snow Making: %s \n" $SNOW
printf "Snow Life: %s \n" $LIFE
printf "\n"



You invoke the script by feeding it the four letter code of an NWS observation station (for instance KEKN for Snowshoe or KSHD for Wintergreen). A list of NWS stations and codes is at:

http://nws.noaa.gov/data/current_obs/

Craig
Scott - DCSki Editor
January 13, 2005
Member since 10/10/1999 🔗
1,252 posts
Craig,

I like your choice of platform. (Note that DCSki runs on Mac OS X on an XServe.)

- Scott
skier219
January 13, 2005
Member since 01/8/2005 🔗
1,318 posts
Quote:

Craig,

I like your choice of platform. (Note that DCSki runs on Mac OS X on an XServe.)

- Scott





Hey, Cool! We have a rack of Xserves at work and they are good machines. OS X is the nicest UNIX OS I have worked with -- it's always my first choice for software development.

Craig

Ski and Tell

Snowcat got your tongue?

Join the conversation by logging in.

Don't have an account? Create one here.

0.15 seconds