Predicting cloud top elevations

In another thread, androolus wrote:

This is more a rant about weather forecasting lameness than a trip report. It was quite sunny at Muir on Monday. We had a delightful tour though it started in the fog and ended in the fog. The fog creeped up to 7500 by 3pm when we were descending, making the nisqually chutes a little more dicey. But clearly there was overnight freezing and the corn was pretty good from muir to the bottom of the chutes. I was expecting it to be too soft for good skiing but I was pleasantly surprised.

I saw pictures from rainier, adams and st helens showing nice clear skies (above a certain elevation)...so I'm wishing that the NOAA forecasts could have indicated clearing above a certain elevation...and that I hadn't cancelled my plans for an adams fest. Did anyone see a forecast showing clearing above?

The Weather Service generally doesn't include cloud top elevations in their forecasts, probably because there's a tiny market for this and it's only occasionally practical to forecast the cloud tops. You have to infer the cloud-top elevation from other aspects of the forecast. Make sure to read the forecast discussion:

www.wrh.noaa.gov/forecasts/display_speci....php?sid=SEW&pil=AFD

When the primary weather pattern is "onshore flow" (like today) it implies a moist layer of marine air that is not associated with a weather front. In this pattern, the cloud tops typically have a uniform elevation. You can infer something about their thickness based on what time of day the Weather Service expects clearing to occur. When no afternoon clearing is predicted (like this Memorial weekend) the clouds are generally quite thick (like 6000ft+ tops).

The best source for cloud-level information is the NOAA aviation weather service:

www.aviationweather.gov/adds/

The only good way to get cloud-top information is via aircraft pilot reports (PIREPs). I don't think the NOAA aviation service publishes cloud-top elevations based on a forecast (not reliable enough.) They only publish cloud-top elevations based on actual pilot observations. Here is a direct link to the PIREPs page:

www.aviationweather.gov/adds/pireps

You can look up the abbreviations for airports here:

www.webflyer.com/travel/milemarker/lookup.shtml

The abbreviation for Sea/Tac is "SEA". In the PIREPs page, if you plug in "SEA" for the airport code and 36 hours for the time period, you'll get something like this:

---

Output produced by PIREPs form (1620 UTC 26 May 2015)
found at www.aviationweather.gov/index.php

Raw text PIREPs

VR UA /OV CYWH 240015 /TM 1705 /FLUNKN /TP B105 /SK 010-012OVC /RM VC SOOKE
KLS UA /OV KLS/TM 1715/FL018/TP BE35/SK OVC018-TOP026/RM DURC EAST -ZSE
OLM UA /OV OLM160030/TM 1720/FL018/TP C182/SK OVC018-TOP035/RM DURC SOUTH -ZSE
HQM UA /OV HQM102010/TM 1906/FL050/TP RV7/SK TOP040/IC NEG/RM -ZSE
AST UA /OV AST/TM 1918/FL015/TP P28A/SK VRB015 TOP039/RM -ZSE
AST UA /OV AST/TM 2055/FL020/TP C680/SK OVC020-TOP047/RM -ZSE
AWO UA /OV SEA007055/TM 2250/FL260/TP E170/TB LGT-MOD CHOP 260-220/RM -ZSE
SEA UA /OV SEA/TM 0215/FL045/TP B737/SK OVC022-TOP045/WV 22005KT
AST UA /OV AST/TM 1410/FL024/TP PA31/SK BASES OVC 024 TOPS 037/RM -ZSE

---

PIREPs are recorded using Universal Coordinated Time (UTC) which is 7 hours ahead of Pacific Daylight Time. The TM field indicates the report time. (Use google to find pages that describe how to read PIREPs.)

Down near the bottom there is a report for Sea/Tac (SEA) recorded at TM 0215, which means 7:15 pm PDT on May 25, I think. Here's the full report:

SEA UA /OV SEA/TM 0215/FL045/TP B737/SK OVC022-TOP045/WV 22005KT

Here's the important information about cloud conditions:

SK OVC022-TOP045

This indicates sky (SK) conditions were overcast (OVC) at 2200 feet with cloud tops (TOP) at 4500 feet. (The elevation is expressed in hundreds of feet.) That means you'd have had a decent chance of getting above the clouds at Paradise at that time.

If you run the same PIREPs report by selecting "Plot" for the output, you'll get something like this:



This is helpful to get a feel for the geographic distribution of the conditions, but be careful if you enter a long time span like 36 hours, because you'll see a mix of reports spread over time. Note that many stations don't report cloud tops at all, probably because they don't have current pilot observations.

With this information and the overall weather forecast you can make guesses about what the cloud layering will be, based on the forecast and what the layers were like yesterday. This can be tricky when the layering is changing from day-to-day (like it did this weekend). You have to roll the dice a little.

Another way to infer the cloud structure is via visible satellite loops. I have this one on my weather page :

www.atmos.washington.edu/~ovens/loops/wxloop.cgi?vis1km+12

As I write this, I can actually see Mt Rainier poking above the clouds on this loop. But that's rare. You have to infer cloud thickness by where the edges of the clouds are and whether they are increasing or decreasing over time.

Very cool, thanks for writing this up.

There is also the cross section plot for Rainier on the UW website (www.atmos.uw.edu/~ovens/wxloop.cgi?mm5d3_cx8+///3). It shows wind vectors (green), relative humidity (red contour lines), potential temperature in Kelvin, and cloud mixing ratio (coloring of the plot). If I see any color (even blue), I assume white-out conditions. This may be too conservative, but reflects my experience with the area cloud plots at 12km that Lowel mentions above. I also look closely at relative humidity.

Over the past week I have been looking at the cross section plot more closely, but it seems to have been hit or miss. I asked a friend who is a pilot to take pictures of the top of Rainier for me whenever he flies by it. Right now the model doesn't seem to match what I am seeing in the pictures. Note though that I have only limited samples. Maybe it's just time shifted. I need to collect more data to have better ground truth. If the model is off, I am wondering if the easterly pattern and the large vertical convection (clouds up to 40k+ feet/thunderstorms) are making modelling difficult for a feature like Rainier.

The other way to correlate might be to look at skewed t-log charts for an airport near Rainier.

Lowel is correct that most weather forecast products do not include cloud tops. There is one exception, the Aviation Area forecast does contain cloud layering and tops. Unfortunately these are hard to read for non aviation people. Look for "top". Clouds may be layered and you may have multiple tops in there...

www.aviationweather.gov/areafcst/data?region=sfo

Here is the Cascade section.

[...]
SYNOPSIS VALID UNTIL 271400
CLDS/WX VALID UNTIL 270800...OTLK VALID 270800-271400
[...]
SYNOPSIS...ALF..BROAD UPR LVL TROF PAC NW/ID/NWRN HLF MT MOVG LTL
THRU PD. STNR UPR RDG ERN PAC WTRS ALG 130W.  WK UPR FLOW DURG
PD.
SFC..STAGNANT SFC PATTERN WITH NO FNTL SYSTEMS XPCD THRU 14Z.
[...]

WA CASCDS WWD
CSTLN...BKN020 TOP 030. 00Z SCT030. BECMG 0608 BKN015 TOP 025.
OTLK...MVFR CIG.
INLAND
STRAIT OF JUAN DE FUCA...SCT020 SCT060. 03Z SCT030 BKN100 TOP
170. OTLK...VFR...12Z MVFR CIG.
PUGET SOUND...BKN030 TOP 060. 22Z SCT040. BECMG 0305 BKN080 TOP
170. OTLK...VFR...12Z MVFR CIG.
S INTR...BKN030 TOP 060. 23Z SCT040. BECMG 0507 SCT040 BKN090 TOP
170. OTLK...VFR...10Z MVFR CIG.

And a translation into verbatim to the best of my knowledge. Once you know a few items it isn't super hard.

Synopsis valid until 05/27 7AM (14:00 Zulu)
Clouds/weather valid util 05/27 1AM... outlook valid between 05/27 1AM - 7AM
[...]
Synopsis.... aloft ... broad upper level trough over pacific north-west/ID/northwestern half MT moving little throughout period. Stationary upper ridge earn pacific waters along 130W. Weak upper flow during period.
[...]

WA cascades westward
Coastline .... broken 2000ft, cloud top 3000ft, 05/26 5PM (00Zulu) scattered 3000ft. Becoming between 11PM-1AM broken 1500ft top 2500ft. Outlook MVFR ceilings (marginal visual flight rules; < 3000ft ceiling, 3-5mi visibility).
Inland
Straight of Juan de Fuca... scattered 2000ft, scattered 6000ft, 05/26 8PM scattered 3000ft, broken 10000ft, top 17,000ft. Outlook ....VFR (visual flight rule; > 3000ft ceiling, > 5mi visibility). 05/27 5AM MVFR ceilings.
Puget Sound ... broken 3000ft, top 6000ft, 3PM scattered 4000ft. Becoming 8PM-10PM  broken 8000ft top 17,000ft, outlook...VFR... 05/27 5AM mvfr ceilings.
Southern interior... broken 3000ft, top 6000ft, 4PM scattered 4000ft. becoming 10PM to midnight scattered 4000ft broken 9000ft top 17,000ft. outlook...vfr....05/27 3AM mvfr ceilings.
[...]

There's also this - click . I use it to determine the tops of inversions that we get pretty frequently in the winter on the east side of the mountains and I find it to be quite accurate. Enter the three letter abbreviation for a particular airport (I use Omak (OMK) and if there's a definable top to the cloud deck the red and blue lines will diverge laterally and noticeably. It's also good for temperatures and windspeeds at various elevations.  

Ah- I see that Schwerkraft already referred to this chart above. Not sure how useful it would be on the west side but it's sometimes useful over here..

There's also this - click . I use it to determine the tops of inversions that we get pretty frequently in the winter on the east side of the mountains and I find it to be quite accurate. Enter the three letter abbreviation for a particular airport (I use Omak (OMK) and if there's a definable top to the cloud deck the red and blue lines will diverge laterally and noticeably. It's also good for temperatures and windspeeds at various elevations.  

I was about to post this, but Matt mostly covered it. The first pictures shows how to set it up. In my case KOLM (Olympia). The second picture shows the skewed t-log plot. I believe analysis in the title implies that this was the actual data taken of a weather balloon. If you hover over the graph you can see the elevation and a bunch of other data. As Matt points out the converged red and blue line tell you where dew point and temp are converging and relative humidity gets to near 100%. hence clouds Where they go sharply apart shows the top.

On the setup page you can ask for a forecast as well. Then the title will say forecast.

And then there is looking out the window aka Muir web cam. Unfortunately not yet up for the season.

www.nps.gov/mora/learn/photosmultimedia/webcams.htm

Getting a good estimate of forecast cloud-top heights is fairly easy using the regional MM5 model from the University of Washington. Follow this link to the forecast soundings site.  www.atmos.washington.edu/mm5rt/rt/soundi....cgi?GFS+current_gfs

Mouse over to a site close to your desired  trip location.

You will then see a forecast “sounding” for every 3 hours out to 84 hours. The horizontal bar near the top of the page gives buttons for forecast time. The forecast time for the displayed data is given in PDT. The sounding is displayed graphically, and in a table. In the graph, the Y-axis is the the height in a pressure coordinate (Mb), and the X-axis is temperature in degrees C. There are two curves in the graph. The left hand curve is the dewpoint or a measure of humidity as a function of altitude.  The right hand curve is the temperature as a function of altitude. When the two curves are close together, the air is humid or cloudy. When the curves are far apart, the air is dry or clear.

The table displays the same data. Look for elevations where the temperature and dewpoint are within a degree or two of each other. At those locations, the air is likely mostly cloudy or cloudy. The table gives heights in meters. The graph and table are a nice way to get cloud base and cloud tops as a function of time and location. Also available is a forecast of winds as a function of altitude.

Remember, these are forecasts from a high resolution regional model and are generally pretty accurate.  As you step out in time, forecast confidence decreases.

The marine cloudiness is prevalent in later spring in the NW. To get a good idea of the likelihood of a “marine push”, check the pressure gradient between North Bend, OR and Seattle (OTH-SEA) late in the afternoon. (www.atmos.washington.edu/marka/pd.cgi). If it is greater than 4 mb, a marine push is likely. Expect cloud tops between 6000 – 8000 ft. If the pressure difference between Seattle and Yakima (SEA-YKM) is greater than 4 mb, expect windy conditions on the east slopes of the Cascades.

Bryan

Thanks to all the folks contributing here. This thread is going to up my cloud forecasting game for sure....