/GMSL Data Correction: Follow-up

GMSL Data Correction: Follow-up

Brief Note by Kip Hansen — 12 February 2020

 

FEATURED_IMAGE


One month ago I posted a
Brief Note here about a correction to the GMSL data and graphics at NOAA’s
Laboratory for Satellite Altimetry (LSA).   The problem at that time concerned the data files and their associated graphics, often used here at WUWT and elsewhere as definitive and reliable up-to-date information of Global and Regional Sea Level Rise, found from this web page:

 

 

star_nesdis


This is the link:

https://www.star.nesdis.noaa.gov/sod/lsa/SeaLevelRise/LSA_SLR_timeseries_global.php

At the bottom of the page illustrated, you see four columns of links to various forms of the data.  The data is noted to be: “Only altimetry measurements between 66°S and 66°N have been processed. An inverted barometer has been applied to the time series. The estimates of sea level rise do not include glacial isostatic adjustment effects on the geoid, which are modeled to be +0.2 to +0.5 mm/year when globally averaged.”

The different versions of the data are:

four_data_formats

Last January, there was a glitch in the data and the graphics that resulted in 2 data points from the month of December to be missing from the files and images. They looked like this:

missing_data

Up at the upper right hand corner, you can see the one latest data point, and before it, blank space where at least two other data points should have been.

I wrote to the Webmaster of STAR, Eric Leuliette, who was helpful and replied:

“For 2 cycles of Jason-3 data, our database was missing some data due to a script failure. I’ve fixed the database and rerun the mean sea level data. The replacement files are on our web site now.”

Unfortunately, it appeared that the fix broke something else, and the result of the repair was this:

csv_and_png_disagree

It appears that the missing data points for December have been added in but according to the data file for this graphic, the point you see above at the intersection of the red lines, the first data point for January 2020,  is shown to be just under 70 mm.  The data file says it should be 63.41.

Naturally, I politely notified Eric of the new discrepancy by return email….and never heard back.

FAST FORWARD ONE MONTH, TO TODAY…..SITUATION:  It is Worse Than We Thought

Here is the image of the same data set pictured above, download at 1400 hrs 11 February 2020.

11_Feb_global_free_all

Oops, it is exactly the same at it was on 10 January, no new data points have been added, there is no correction for the apparently missing/incorrect data for 2020.022.

In fact, checking all four of the graphics, and all four of the data files (.csv files), NONE of them have been updated since 2020.022 (one on 2020.023).   It is not only the global mean sea level data sets and graphics, clicking through to the Regional sea level time series page, we find the exact same problem — nothing has been update since the first data point in January 2020.  The ninth of February is, in digital date format, 2020.040.

For the year 2019, the .csv (comma-separated values) file shows 36 entries for the year, one approximately every 10 days.  When investigating the previous problem, there were two missing values, each about ten days apart.  This year, the last date for which there is any entry is, any of the sea level rise files (or on their graphics) is 2020.022 (in one case, 2020.023, which represents the same date).

The Official Scoop:

The webmaster at NOAA STAR/NESDIS, Eric Leuliette has been very responsive — the epitome of a public employee fulfilling his responsibilities to the general public.   Truthfully, he has been more than patient with my several emails and many questions, responding and clearing up a less-than-simple situation.

Point 1:   Why have these images and files not been updated since 9 January 2020?

Answer:  Jason-3 has been in safehold this week, so it’s likely that there will be missing data for one or two cycles.”   

and

CNES is investigating the source of the Jason-3 safehold.”

“The safehold position safeguards instruments by drastically reducing the satellite’s power usage, but it has the perturbing side effect of forcing DSCOVR [in this case, Jason-3 — kh] to stop transmitting data back to Earth.” [ source ]

Point 2:  Why are the data files (.csv files) different from the graphics files which share the same name?

noaa_green


Answer:  “
The plots have been smoothed with a 4-month boxcar filter. I have changed the text on the website to note this. This is the source of the differences that you noticed between the csv files and the plots. The attached figure shows the smoothed values (green) versus the unsmoothed values (red).”

As promised, the text on LSA/Sea Level Rise page has been changed to read:

“Only altimetry measurements between 66°S and 66°N have been processed. An inverted barometer has been applied to the time series. The estimates of sea level rise do not include glacial isostatic adjustment effects on the geoid, which are modeled to be +0.2 to +0.5 mm/year when globally averaged. The plots have been smoothed with a 4-month boxcar filter.” [the bolded text was added –kh]

Point 3:  Why don’t the data files (the .csv files) agree with the plots?

Answer:  “The plots have been smoothed with a 4-month boxcar filter. I have changed the text on the website to note this. This is the source of the differences that you noticed between the csv files and the plots.”

Point 4:  What?  Where are the data files with the smoothed data?

Answer:   “We have not been providing a file with the smoothed values.”

Explanation:  The data files contain data points for each “10-day pass” of the Jason-3 satellite.  Apparently, the script that produces the images (.png files) shown above, the multi-color, multi-satellite graphs, first runs a 4-month boxcar filter on the data and then puts the smoothed data on the graph — not the original data.  This is true, even though the .csv file and the graphics share the same name — they do not share the same data points.

Point 5:  When we will see updated data and images?

Answer:  “I only update the time series sometime around the 15th of each month.”

And, as Jason-3 is/was in safehold, it will depend.

 # # # # #

 Author’s Comment:

 I promised last month that I’d let you know if there was any progress on this, expecting an answer in a few days.  It took longer but now we know.

I normally wouldn’t fuss with an oddity like this, but I write about Sea level Rise and the STAR/NESDIS sites are my mainstay for up to the minute SLR data from the Jason series satellites.

I have been very impressed with NOAA’s response — from Eric Leuliette, the Center for Satellite Applications and Research (STAR) webmaster, who has gone the extra mile to answer my questions.

# # # # #