DST product is useless(nearly)

[Christopher S.]

Hello. I'm here to share this

 

There has been a notable and fairly consistent delta between the predicted and the observed "disturbances" which is both confusing to look at and scientifically not viable. I believe this community worked out a few years ago that one of the primary stations used to produce DST data went down, and ever since, there has been this large delta between the two that I will not continue to ignore. I don't know what you could possibly do about this, but I am bringing it to your attention now because just look at it.

Edited May 6, 2023 by Christopher S.

[MinYoongi]

It was like this even before the station went down. The station being down just caused the real data to be more negative than it really is by like -50NT. The prediction sadly was always crap. I dont know how they calculate it, its kinda like the GFZ KP prediction.

 

Edited May 6, 2023 by Sam Warfel
Duplicated text

[Christopher S.]

8 hours ago, MinYoongi said:

It was like this even before the station went down. The station being down just caused the real data to be more negative than it really is by like -50NT. The prediction sadly was always crap. I dont know how they calculate it, its kinda like the GFZ KP prediction.

 

I'm not sure where the delta came from, ultimately. The predicted values on average, when there's really not much activity, have a delta of about -20Nt to the observed, and during activity this difference is further exacerbated. This usually betrays the timing of storms subsiding, and could very easily be fixed with a +20Nt offset to the predicted values. At the end of the day it seems like an offset was placed on accident at some point, so maybe just removing that offset. As far as what SWL could do, I don't know how ethically challenging it would be to change this product and present a more accurate version of the DST, but that's what I'd do. Just add +20Nt offset.

[Sunshine]

The ethics involved would be the science ethics. I have to say, I'm from a time, when consensus wasn't enforced and a healthy debate about methods and results was encouraged. I have no idea what would be involved in changing the DST calculation here on this website, even though my day job is in software development. (embedded, biotech stuff)
From the scientific point, I personally would believe, that with creating better prediction tools and publishing them to the scientific community, said community would be very happy. As long as the algorithms are published, the inputs and outputs can be verified, I don't see anything wrong with deviating from the what is being used now.