Snow Watch 2022

It looks like it might snow in the Berg this weekend. @JonWells, any expert forecasts?

 


The Icon model is holding on to some snow this weekend. 

it’s unusually far north - normally the snow hits the southern Berg more. 

i think this is because the mechanism that is making the snow is a cut off low produced by a breaking Rossby wave in the upper atmosphere which leaves behind a blob of very cold Southern Ocean air. That blob is thrown quite far north by the breaking wave. In the case of a breaking wave, the cold air becomes semi-stationary over southern Africa for a few days. If the wave does not break (i.e. no cut off low) then the wave moves off to the east in a day and the mechanism making the snow doesn't get the chance to dump on the high ground of Lesotho - it will melt in the atmosphere before it hits the ground somewhere to the east of Lesotho and fall as drizzle.

 

Winds in upper air-
The blue colour over the SA/Botswana border is the middle of the cold blob of air that has broken away from the Southern Ocean air.

 

The Icon forecast model held on to the localised blob of snow in the far northern Berg right up to real time (see below)

This seems to be because that is where the model was coldest (see forecast model temperature map above).




The ECMWF model simulated much more widespread snow across the Berg instead (see below). 

Snow has been reported at Sani with a dusting evident on the Cathedral Peak range. 

JonWells commented on the large spread in the forecast models at the start of the medium range forecasts. That large spread was retained up to real time. Forecast models are initialised by real measurements of the atmosphere. What the models forecast into the future depends more and more on the forecast model the further ahead in time the model runs and the further away they are in time from the real starting conditions. But as the event of interest comes closer in time, the models should converge to a similar forecast. If this doesn’t happen - as was the case here - then it is because the observations of the real atmosphere are too sparse i.e. too far apart from one another or else non-existent. In the case of southern Africa, observations are critically sparse so the models are not brought precisely to heel as the event of interest approaches. And the view the models have of the world is their own view, not that of the real world.  All models predicted the severe event this weekend but differed in detail. Where that detail depends on the models getting a key threshold like the altitude of freezing point just right - as is needed for snow forecasts - the differences between the models are much easier see. 

The question for those in the mountains is whether one model is actually more reliable than another. The modelling centres that build the models will evaluate the models for a wide range of metrics, usually those of most interest to the country where the modelling centre is located. For something like snow in the Berg, with only a handful of us interested, that evaluation won’t happen. And since model improvement requires those metrics as a starter, we’re not likely to see deliberate improvements. Neither will we see a return to less sparse conventional observations over southern Africa such as weather balloons. That network over southern Africa was at least 10 times better in the 1970s than it is now. The network is in terminal decline. A bit like Eskom. 

 
  
 

During this past week I did a three day hike at Giant’s Castle of the Contour Path from Langalibalele Ridge to Taylor’s Pass, entering via Langalibalele Ridge and exiting via Oribi Ridge Route, and although I wasn’t in the snow I got a good view of the snow on the summit during my hike lower down.

The snow on Giant’s Castle, Giant’s Castle Pass and Long Wall from Langalibalele Ridge on Tuesday.  

 

The Carbineers peaks and top of Langalibalele Pass from the contour path.  

 

I had decided to include the smuggler’s shortcut over the ridge into Elandshoek Pass approach into my hike, and to climb to the summit of The Sisters (north) peak while I was on the top of the ridge.

The snow on Elandshoek Pass (the gully to the left of the photo) and South Ridge from the summit of the north Sister on Wednesday.  

 

From the normal approach to Elandshoek Pass, the snow on the escarpment around Redi, The Hawk and The Tent, with snow visible on the southern berg at the left in the background.  

 

On my walk out on Thursday, snow on the escarpment from Langalibalele Pass to Cathkin Peak, the photo taken from Oribi Ridge Route.  



It was apparent that this was quite a significant and widespread snowfall, and more than expected, although the snow was confined to the upper levels of the escarpment and Lesotho, and that the snow on the summit was quite deep.

Can anyone that has been in the high berg between Ndumeni and Champagne Castle within the last couple of days to a week enlighten me on the snow cover on top? Need to pick shoes and I am not too keen on my trail running shoes in snow deeper than 3-4cm in them. This could of course change with new snowfall before Friday but at the moment that looks unlikely.