@cathoderay Are you dissing the Shipping Forecast ? It is only for sailors, surely nobody else understands it anyway. 🤣 .

Posted by: @jancold

↑

It is only for sailors, surely nobody else understands it anyway.

Unless I happen to be a sailor!

The other reason for using the shipping forecast (actually the Inshore Waters Forecast) is that, compared to the normal land forecasts, which are full of waffle, including coverage of weather that has already happened, the shipping forecasts are in a standard format, and relatively concise, and they use key words eg if today's 1200 forecast says SW 6 veering NW 4 later, the key word later means the NW 6 if forecast to happen 12 hours or more after the forecast was issued ie from midnight tonight until 1200 tomorrow (the numbers are the wind speed on the Beaufort Scale, veering means clockwise around the compass). If I want to head SW, I may do better to wait until tomorrow morning, rather than starting out this afternoon.

Even though it is relatively concise, translating the text into something a computer can use is still a challenge. Take the above forecast, that can be coded as two 6 hour blocks of SW 6, then two 6 hour blocks of NW 4, but how about 'S 3 becoming cyclonic 4 for a time'? Cyclonic is shorthand for the typical wind pattern observed while a low passes over, but how do you code 'for a time'? 

I've only found one paper in the literature that tries to compare forecast weather with actual weather. It was done by the Met Office in 2013, and as it happens, it was done using the Shipping Forecasts, for much the same reasons. They stumbled over the actual weather though, the problem being relatively few consistent weather observation posts, and they gave up, and used instead a 'nowcast model', and wrote a classic example of 'no need to bother with real data round here, let's use a model instead' waffle (even if it does include the rather sinister concept of 'truth data'):

"Ideally, real observations should be used to verify forecasts. However, real observations can only be used if they are relatively equally spaced and their coverage is sufficient to represent all the wind speeds within the geographical domain of the forecast. Unfortunately, this is rarely the case so the AFVS [Automatic Forecast Verification System] uses nowcast model analyses as the truth data because they contain the latest observations and the most recent model data." [emphasis added]

Using their version of the 'truth data', they managed to claim the main shipping forecast was accurate within plus or minus one Beaufort scale 88% of the time, while the inshore waters forecasts were accurate within plus or minus one Beaufort scale 94% of the time. Based on casual observations of the real 'truth data', these estimates seem on the high side, perhaps an example of an organisation marking its own homework. If you use a criteria of correct Beaufort scale prediction for the inshore waters forecast, rather than with plus or minus one Beaufort scale, the accuracy drops to around 60-70%, ie the forecast is only accurate two thirds of the time.

Another possible angle on the accuracy is to compare yesterday's inshore waters forecast 24 hour outlook (covers 24-48 hours) with today's actual forecast (covers 0-24 hours). By and large, they should be the same, but if they differ, then at least one of them is wrong. I have yet to try this particular analysis.

All of this does have some relevance to home heating control systems that make use of forecast weather rather than actual weather data to control the heating system. If the actual weather varies from the forecast weather, then the control system will be out of step with the actual conditions, to the extent the actual weather varies from the forecast weather.

Edit: link to the Met Office paper (pdf): Verification of marine forecasts using an objective area forecast verification system.

@cathoderay Don’t know about these days but, Southern Tv had an office in their Southampton premises for the weather forecaster; on the door was a hook, over which hung a piece of seaweed… Regards, Toodles.

Posted by: @toodles

↑

on the door was a hook, over which hung a piece of seaweed

and the curtains were drawn, to prevent the forecaster being distracted by the actual weather... 

@cathoderay I find it very irritating that when they waste some of the brief amount of allotted time on telling us what weather we have been having, they then have to rush through the forecast.😒 I don’t think I need them to tell me what weather I have been experiencing really! Harrumph Toodles.

Posted by: @toodles

↑

I find it very irritating that when they waste some of the brief amount of allotted time on telling us what weather we have been having, they then have to rush through the forecast

I agree completely. Typical of the BBC. The supposed forecast also goes by different names these days, eg the weather situation, or sometimes the ongoing weather situation. Since when was the weather not an ongoing situation? The announcer also needs to read Clive James's Mr Fish's ongoing freezing fog situation report, or situation. Soon the Today pongos will be asking the forecasters for their sense of the ongoing weather situation where they are. God help us all!  

@cathoderay Thanks for the informative reply to my flippant remark. I have been following the forum for a while now and although most goes over my head I hope gradually, bit by bit, to reach an understanding. In the near future I will be going to discuss a design for an ASHP for my house so the more I know the better. The forecast today for Sidmouth said rain stopping by 10:00 , it has not and the "prediction" changed radically just now. There are so many apps that forecast marine weather but they rarely agree with each other. Anyway I am going way off topic.

Posted by: @cathoderay

↑

"Ideally, real observations should be used to verify forecasts. However, real observations can only be used if they are relatively equally spaced and their coverage is sufficient to represent all the wind speeds within the geographical domain of the forecast. Unfortunately, this is rarely the case so the AFVS [Automatic Forecast Verification System] uses nowcast model analyses as the truth data because they contain the latest observations and the most recent model data." [emphasis added]

I know this is probably a silly question but isn’t a forecast based on observed data to start with? i.e. data from the various weather monitoring stations? If so, why can they not use those same observations to compare against the forecast from x number of days ago?

That said, I seem to remember reading that percieved accuracy is based on the predicted weather for now vs the actual weather for now whilst in fact the prediction could be remarkably astute but an hour or two out (and therefore still an accurate forecast within reasonable time variations) and therefore appearing bad. I hasten to add that I am in no way an expert so am happy to be disabused of this viewpoint by anyone prepared to debunk it convincingly.

This topic has taken an amusing twist 🤣

I looked into weather models last year because I found it annoying how, for example, the BBC weather website would have one forecast, and the BBC weather report on TV would have another and it turns they all use different modelling data (not sure why they do that).

What I was able to establish was that weather models are essentially superpowered physics equations running on massive computers. They take in real-time data on temperature, pressure, humidity, etc., then crunch those numbers alongside complex equations describing how air, moisture and energy move. This creates a simulated atmosphere, but the sheer number of variables and limitations in starting data can lead to errors that grow over time, making long-term forecasts less reliable. One analogy I came across was comparing weather prediction to trying to predict the exact path of a butterfly flapping its wings days from now. Weather models grapple with a similar kind of chaos, though on a much larger scale.

@editor Some of the most powerful supercomputers have been built and installed for meteorological prediction systems. ECMWF have their centre just a mile or so from my heat pump and a French family who are friends of ours have recently retired from using and maintaining one of these beasts, Didier in particular has been the butt of many jokes about the megabucks these things cost and how, as soon as a new model has been installed and commissioned, they are looking at a newer, faster, flashier replacement for it! 😄 Regards, Toodles.

Posted by: @toodles

↑

Didier in particular has been the butt of many jokes about the megabucks these things cost and how, as soon as a new model has been installed and commissioned, they are looking at a newer, faster, flashier replacement for it

In fairness, your average gaming teenager will be functioning in much the same way….

@editor

These are the computer weather models I can choose from in Predict Wind, my sailing weather app. They can be startlingly different so I look to see what suits and then a look out of the porthole and make my best guess. 😎