Support the Arctic Sea Ice Forum and Blog

Author Topic: Erroneous & missing PIOMAS data  (Read 468 times)

Anthony J

  • NewMembers
  • New ice
  • Posts: 1
    • View Profile
  • Liked: 0
  • Likes Given: 9
Erroneous & missing PIOMAS data
« on: May 10, 2021, 07:15:41 AM »
I am just an individual with no vast experience in any of the cryosphere and certainly no contacts within the scientific community, thus I need a little assistance please.

For some years I’ve been using the Arctic Ice Data from “http://psc.apl.uw.edu/research/projects/arctic-sea-ice-volume-anomaly/data/” where I have been picking up the Daily Ice Volume and Daily Ice Thickness Data along with the Average Monthly Ice Volume Data. I believed that all was going well until the end of last year when the Daily Ice Volume Data for 2020 stopped at 365 days to be followed by day 1 of 2021, at the same time the Ice Thickness Data just stopped at day 365 of 2020 and no further data has been filed. So I have checked back at other leap years and the same situation exists for each leap year in that there are only 365 days recorded.

Obviously this leaves me with a lot of questions!!! Like:- Just what day data is missing February the 29th, December the 31st or has this all rolled up and all the data is now out by 11 days? and this applies to both the Daily Ice Thickness and Volume Data. Like:- Just what has happened to the Daily Thickness Data which has just stopped. Like:- Is ALL the Monthly Ice Volume Data incorrect since January 1980 or is it only the leap years that are wrong February onwards?

I have written to the only associated email address I can find, however it is a big bucket address and I have no way of knowing if it is even monitored or what they would do with such a questioning enquiry. So far there has been no response!

What I feel I need is an email address of somebody who would be able to influence the situation or have the historic data corrected and ensure that this is not repeated on future leap years (just in case we get that far!). Either that and/or a web address where I can pick up this information where these errors are not an issue.

If anyone can help me out with this I would much appreciate it. Thank you for your consideration.

Tony

Neven

  • Administrator
  • First-year ice
  • Posts: 7916
    • View Profile
    • Arctic Sea Ice Blog
  • Liked: 1149
  • Likes Given: 572
Re: Erroneous & missing PIOMAS data
« Reply #1 on: May 10, 2021, 10:06:09 AM »
From the PIOMAS page you link to:

Quote
Leap years in PIOMAS are treated by integrating forward for 365 days. For leap years this means that no integration is performed for the last day of the year. The data presented here show the first 365 days of the year.
Il faut comparer, comparer, comparer, et cultiver notre jardin

gerontocrat

  • Multi-year ice
  • Posts: 12832
    • View Profile
  • Liked: 4410
  • Likes Given: 34
Re: Erroneous & missing PIOMAS data
« Reply #2 on: May 10, 2021, 11:30:28 AM »
From the PIOMAS page you link to:

Quote
Leap years in PIOMAS are treated by integrating forward for 365 days. For leap years this means that no integration is performed for the last day of the year. The data presented here show the first 365 days of the year.
For all my sea ice extent and area analyses I decided just to delete the 29th February line from all the data tables, thus making a 365 day year for all years.

This allows the analyses formulae to be standard for all years, and the difference is marginal, i.e. so small as to be statistically insignificant.

purity vs practicality
"Para a Causa do Povo a Luta Continua!"
"And that's all I'm going to say about that". Forrest Gump
"Damn, I wanted to see what happened next" (Epitaph)

Richard Rathbone

  • Nilas ice
  • Posts: 1133
    • View Profile
  • Liked: 236
  • Likes Given: 14
Re: Erroneous & missing PIOMAS data
« Reply #3 on: May 10, 2021, 11:59:18 AM »
Zhang would be the person to contact. http://psc.apl.uw.edu/people/investigators/jinlun-zhang/

Maybe its forced on them because one of their inputs takes a similar attitude to Gerontocrat ;) and they only have 365 days worth of input to work with, but it does seem a bit inelegant not to do the integration on Day 366 even if it only was going to be used for the input to Day 1.


oren

  • Moderator
  • First-year ice
  • Posts: 7598
    • View Profile
  • Liked: 2972
  • Likes Given: 2909
Re: Erroneous & missing PIOMAS data
« Reply #4 on: May 10, 2021, 12:33:31 PM »
Welcome, Anthony J!

A search in the PIOMAS thread for "leap year' finds a couple of posts from two of our most knowledgeable members:

(UW drops day 366 of leap years)

BTW. I fixed the animation for February as well. I was wrong by one year. The excuse is that the piomas files are sometimes confusing at the moment caused by the leap year. There are four months now with the fourth month apparently only one day in it.  I (have to) do the same, sometimes ignoring the leap day and sometimes not.

So it seems this is a known and old issue. I recall that Wipneus compensates for this in some of his processed PIOMAS data files by adding .25 to some days in some years, so I think he catches up to the leap days manually over several years. But can't say for sure.

And Writing Dr. Zhang is certainly an option. He is very nice and responsive, from my own experience.

SimonF92

  • Grease ice
  • Posts: 502
    • View Profile
  • Liked: 199
  • Likes Given: 84
Re: Erroneous & missing PIOMAS data
« Reply #5 on: May 10, 2021, 12:39:39 PM »
From the PIOMAS page you link to:

Quote
Leap years in PIOMAS are treated by integrating forward for 365 days. For leap years this means that no integration is performed for the last day of the year. The data presented here show the first 365 days of the year.
For all my sea ice extent and area analyses I decided just to delete the 29th February line from all the data tables, thus making a 365 day year for all years.

This allows the analyses formulae to be standard for all years, and the difference is marginal, i.e. so small as to be statistically insignificant.

purity vs practicality

Hey, that one row could be the difference between

p=0.04998787

and

p= 0.05119499

And we all know that only one of these is 'real'  ;)
Bunch of small python Arctic Apps:
https://github.com/SimonF92/Arctic

gerontocrat

  • Multi-year ice
  • Posts: 12832
    • View Profile
  • Liked: 4410
  • Likes Given: 34
Re: Erroneous & missing PIOMAS data
« Reply #6 on: May 10, 2021, 01:26:07 PM »
From the PIOMAS page you link to:

Quote
Leap years in PIOMAS are treated by integrating forward for 365 days. For leap years this means that no integration is performed for the last day of the year. The data presented here show the first 365 days of the year.
For all my sea ice extent and area analyses I decided just to delete the 29th February line from all the data tables, thus making a 365 day year for all years.

This allows the analyses formulae to be standard for all years, and the difference is marginal, i.e. so small as to be statistically insignificant.

purity vs practicality

Hey, that one row could be the difference between

p=0.04998787

and

p= 0.05119499

And we all know that only one of these is 'real'  ;)
But I get a day off once every 4 years.
"Para a Causa do Povo a Luta Continua!"
"And that's all I'm going to say about that". Forrest Gump
"Damn, I wanted to see what happened next" (Epitaph)

SimonF92

  • Grease ice
  • Posts: 502
    • View Profile
  • Liked: 199
  • Likes Given: 84
Re: Erroneous & missing PIOMAS data
« Reply #7 on: May 10, 2021, 02:28:13 PM »
gero you are an institution here, im surprised ASIF only lets you have one day off every 4 years, I would personally lobby for 2

Also in the interest of not derailing from the question, welcome Anthony. It looks like PIOMAS do the same thing as gerontocrat and I agree that for a large organisation is it a bit inelegant. They may (probably) have the original data in unix timestamp which would account for this- the stuff they put on the website might be simplified
Bunch of small python Arctic Apps:
https://github.com/SimonF92/Arctic

gerontocrat

  • Multi-year ice
  • Posts: 12832
    • View Profile
  • Liked: 4410
  • Likes Given: 34
Re: Erroneous & missing PIOMAS data
« Reply #8 on: May 10, 2021, 03:09:21 PM »
I agree that for a large organisation is it a bit inelegant. They may (probably) have the original data in unix timestamp which would account for this- the stuff they put on the website might be simplified

a large organisation ? A-team once posted on how he was surprised to find how small it actually is.

The Polar Science Center is a department within the Applied Physics Laboratory at the University of Washington.

The PIOMAS production is really just a couple of guys.

A shoestring operation always searching for funds, methinks ? We are lucky to have it.
"Para a Causa do Povo a Luta Continua!"
"And that's all I'm going to say about that". Forrest Gump
"Damn, I wanted to see what happened next" (Epitaph)

SimonF92

  • Grease ice
  • Posts: 502
    • View Profile
  • Liked: 199
  • Likes Given: 84
Re: Erroneous & missing PIOMAS data
« Reply #9 on: May 10, 2021, 03:21:54 PM »
I stand corrected. Its quite unsettling thinking that one of my main "go to" external sources stands on what probably is uncertain ground/grant-funding in the grand scheme of things
Bunch of small python Arctic Apps:
https://github.com/SimonF92/Arctic