Re: weather forecasts

I know that it's difficult to make exact predictions about weather in the mountains, but I've been flitting around between weather sites, and there seems to be something crazy going on at Weather.com. They claim the high today was 59 degrees. The National Weather Service says that it's 70 degrees there right now. National Weather Service says the lows next week are going to hover around the mid-40s. Weather.com says Yosemite's nights are going to get down to the low 20s.

One of the things to remember about weather forecasts is that they are always approximate. I went to Moose Lake a week ago. The NOAA website map showed a low in the low 20s if you clicked on one end of the lake, and a low in the 30s if you clicked on the other. We went ready for the low 20s and it was in the low 30s.

I would believe the forecast for the 20s for the 7000-8000 elevation range in Yosemite. Last night at around 9000 feet in Kaiser Wilderness, it was in the mid 30s, until early morning when it dropped to the low 20s as per usual (early in the morning it usually dips to the lowest low of the night). I woke to frost. It was around 60 degrees when I reached the car, and it's in the 70s here in the central valley today.

I would anticipate lows in the 30s for Yosemite Valley.

You have to watch the elevation used in the forecast. In mountainous areas the elevation can be very different from the exact point on the map you clicked on. Low temperature forecasts are likely to be more inaccurate than highs because of local temperature inversions and pooling of cold air in low spots.

Quote
Calaveras
You have to watch the elevation used in the forecast. In mountainous areas the elevation can be very different from the exact point on the map you clicked on. Low temperature forecasts are likely to be more inaccurate than highs because of local temperature inversions and pooling of cold air in low spots.

I'm pretty sure the west end of the lake was at the same elevation as the east end. It looked pretty flat and wasn't spilling off the edge of the granite bench.

I don't know exactly what forecast you were using. The gridpoint forecast covers a small square and even though you can click on any point on the map, you only get the forecast for the square. In some situations the ends of a lake can be in different squares using different elevations. Yosemite Valley is like this. You can get anything from 4000' to 8000' depending on where you click. Can you provide a link to the forecast you were using? I'd like to see it.

As I mentioned, the other problem is local inversions and cold air pooling, neither of which can be accounted for by the forecasts. Only detailed knowledge of the exact spot can tell you what to expect.

Many years ago I camped at Bridalveil campground as part of the Glacier Point public astronomy program in August. Glacier Point was pretty balmy at night but I about froze to death at Bridalveil.

I was using NOAA's website, the source of all weather reports.

search up Three Rivers, CA. Using the map, crawl your way east and somewhat north to Moose Lake. Click on one end of the lake, and it will consistently be 10 degrees different than the other end of the lake.

I suspect the one end is predicted using the weather station on the ridge over Pear Lake, and the other end of the lake is sourced from some other weather station. It is too consistently different. The lake is the size of a football field and not in actuality differing in temperature or weather from one end to the other.

Calaveras is right. The forecasts are based upon a grid where each square within the grid is forecasted. That square is given its own elevation. It so happens that Moose Lake is divided between two squares in the grid (check out the green overlay square, its the entire grid square for that forecast. The 'elevation' of the 'grid square' is given just below the map.

Here's the link for the east end of Moose Lake. Note that it is within a 'grid square' which 'has' an elevation of 9368'. Tonight's low 22, tomorrow's high 35.

Here's the link for the west end of Moose Lake. Note its 'grid square' 'has' an elevation of 10978'. Tonight's low 9, tomorrow's high 20.

You're looking at forecasts for 'elevations' which are about 1600' apart.

On the NWS point forecast maps I always check the claimed elevation and, if necessary, click nearby areas outside the 'grid' to see if those locations have an 'elevation' closer to what I know to be the real elevation I'm interested in.



Edited 1 time(s). Last edit at 10/10/2012 06:15PM by ttilley.

The ends of the lake are at the same elevation. I've been there...

I know the lake isn't spilling off to the east.

Imagine a 3-D image of the area. Now imagine you're trying to do a weather forecast for the entire area. The space across which you're forecasting is continuous, so there's an infinite number of possible "locations" you're forecasting. Can't do that, so imagine instead you're forecasting each single square inch, so you assign a location and elevation to each square inch. How many square inches are you forecasting, and do you have enough budget to do that? Assume the answers are "too many" and "no". So you need a larger "square" size. Let's suppose your square is, instead, 2 miles by 2 miles (which by eyeballing the green square which comes up when I request a forecast for my home appears close to the actual square size). Now you have a lot fewer squares to forecast, and it fits within your computing budget.

The problem is...each square must be assigned an elevation, which will apply to the entire square. Not so bad here in the flatlands around SF Bay, but where the terrain is very mountainous your 3-D image of the area is now extremely lumpy, it doesn't quite look like the 3-D image you started with.

When you click on a point in that map you're really just selecting one of these squares. They probably should find a better way of conveying this fact, but that green square is showing you the "square" you selected (the crosshairs is probably a bit deceptive, you didn't really select a specific location even though it looks like you did). Note that the east-side Moose Lake square covers not just the east end of the lake but the deep valley east of the lake, and note that the west-side square covers some high country west of the lake.

But if you're not really paying attention, you click on one, not the other, you get more inaccurate information than you would have. Which is all most people do - glance at the screen, go for it. I had a hiker assuming he would get by with a 40 degree bag cause he's a warm sleeper - reminding him of the 20 degree forecast at the other end of the lake probably increased the margin of safety as he then came prepared for lower temps. (This time of year at higher elevation, the lower rated bags start to go in the pack for me as a default. It no longer matters to me what the forecast actually says.)

So this little trick is noted as a planning consideration, if you use the NOAA map.

Quote
AlmostThere
The ends of the lake are at the same elevation. I've been there...

Most lakes work that way.

Quote
eeek

Quote
AlmostThere
The ends of the lake are at the same elevation. I've been there...

Most lakes work that way.

Where are the lakes not part of this "most" that don't work that way?

Quote
chicagocwright

Quote
eeek

Quote
AlmostThere
The ends of the lake are at the same elevation. I've been there...

Most lakes work that way.

Where are the lakes not part of this "most" that don't work that way?

Well, there is a really big one under the continental ice in Antarctica that is more of a spheroid. The Russians have drilled about 12,000 feet down to it....

http://www.nytimes.com/2012/02/09/world/europe/russian-scientists-bore-into-ancient-antarctic-lake.html

Quote
chicagocwright

Quote
eeek

Quote
AlmostThere
The ends of the lake are at the same elevation. I've been there...

Most lakes work that way.

Where are the lakes not part of this "most" that don't work that way?

Ones big enough to have tides?

The Great Lakes don't have significant tides, but they do get seiches. That would do it.

I like to use what the NWS used to call the Gridpoint Forecast. You can click around a map and get a 7 day forecast. It shows the elevation used for the forecast. The temperature forecast has always been fairly accurate.

Here's a starting point:

Cathedral Lakes Forecast

I agree with Calaveras regarding using the Gridpoint Forecast. I wouldn't vouch too much on its quality, however, having experienced up to (in the worst cases) plus or minus 15F from the forecasted temperature. My personal rule of thumb is to look at the Gridpoint Forecast for the area I'm heading to and prepare for 5-10F lower (I would rather be over-prepared than have a miserable cold night).

On a more objective note, I camped in the vicinity of Ireland Lake (close to 11,000 feet) on Saturday (October 6th) night. I was expecting low 20s and was (pleasantly) surprised to experience low 30s (no frost and no condensation on the tent).

I use intellicast.com. Been using it for a number of years now.
Seems pretty reliable.

My general rule of thumb is simply:
30 degree bag is put away come October and 15 degree is used.
Since always bring lightweight poofy pants and jacket forecast is pretty irrelevant
w/r to temps.

Have fun

---

It's getting cold out there this past week. The past couple of days the weather station in my backyard has been recording lows in the mid-40's. This is near sea level in the Bay Area. So I wouldn't be surprised that in the Sierra it's in the 20's if not lower.

.