Bryan,
Thanks for mentioning the differences in the ONI methodology. I think I understand the situation now.
The problem is that I cannot easily change MGET to obtain the ONI data from the table on the
CPC ensoyears.html page. When I developed the ONI capability, it was part of developing a tool that can read all of the indexes that ESRL PSD lists
here. All of those indices are available there in a common text file format. We had need of ONI in one project and NAO in another. ESRL PSD’s effort was a convenient way to get them both,
plus many more.
Currently, ESRL PSD describes their
ONI index as “Oceanic Nino Index From
NOAA Climate Prediction Center (CPC). Three month running mean of NOAA ERSST.v2 SST anomalies in the Nino 3.4 region (5N-5S, 120-170W), based on the 1971-2000 base period.
Time Series is a newer version from source!” (this is all of their formatting). I looked at the values. They do not match the
CPC ensoyears.html page that their description points to, but nor do they match CPC’s
Link to Previous Version of ONI (1971-2000 climatology).
So, unfortunately, I can’t say
what those values are that ESRL PSD is currently providing. I expected them to at least match CPC’s
Link to Previous Version of ONI (1971-2000 climatology)
but they don’t. My guess is that it is a previous version of CPC’s data that hasn’t been updated since mid-2013.
The easiest solution for me would be for ESRL PSD to make their data current with what CPC provides. But perhaps not even that would be optimal for the future,
as ESRL PSD does not mark the ONI index as “regularly updated”. I find this odd as I’d consider ONI the most widely used index on their entire list. I will contact them about this, CC you guys. If they are not committed to maintaining an updated ONI, I will
contemplate other solutions. Hopefully there would be an alternative to scraping the HTML table from the ensoyears.html page…
Jason
From: Bryan Costa - NOAA Affiliate [mailto:]
Sent: Tuesday, November 04, 2014 11:55 AM
To: Jason Roberts
Cc: ; Arliss Winship - NOAA Affiliate
Subject: Re: [mget-help] Re: Question about ENSO monthly binning
Hi Jason,
To be honest, it would be preferable to have MGET use the more recent ONI table, although a warning would be really great in the interim. The reason that an update would be useful is because ONI values are calculated (by NOAA) against
a different 30 year base period every 5 years. When these 5 year updates occur, the ONI values over the most recent decade will change because of the inclusion of more recent data.
The ONI table that MGET currently points to uses the base period of 1971-2000. The latest table derives the ONI values from 1981-2010. I'm pretty sure this is the reason for the discrepancies that Arliss discovered. The next ONI update
is scheduled for 2016. Please see the following website for more info: http://www.cpc.ncep.noaa.gov/products/analysis_monitoring/ensostuff/ONI_change.shtml
Bryan C.
On Tue, Nov 4, 2014 at 10:56 AM, Jason Roberts <> wrote:
Bryan,
That is a good question. I looked at the code. Your guess is correct. Images that occur outside of
the span of dates covered by the ONI table are silently omitted from the climatology.
Do you think I should modify the tool to report a warning about this?
Jason
From: Bryan Costa - NOAA Affiliate
[mailto:]
Sent: Tuesday, November 04, 2014 10:09 AM
To: Jason Roberts
Cc: ; Arliss Winship - NOAA Affiliate
Subject: Re: [mget-help] Re: Question about ENSO monthly binning
Hi Jason,
I had a quick follow-up question. If MGET uses the ONI table that only goes to March 2013 (click here),
how is MGET classifying products into ENSO cycles acquired after March 2013? Are these products left out?
On Mon, Nov 3, 2014 at 7:47 PM, Arliss Winship - NOAA Affiliate <> wrote:
Hi, Jason.
Thanks very much for the information. That answers my question. I may e-mail the NOAA ESRL PSD folks to ask about the differences I saw. If I do I'll let you know what they say.
Thanks again.
Arliss
On Mon, Nov 3, 2014 at 11:52 AM, Jason Roberts <> wrote:
Bryan, Arliss,
To determine ONI episodes when binning rasters into ENSO phases, MGET uses this logic:
1.
Download this table:
http://www.esrl.noaa.gov/psd/data/correlation/oni.data. This is the same URL that Bryan sent, and is the one that appears on
this page under ONI. However, I see that it only goes through March 2013, while the
ensoyears.html page that Arliss referenced (which I am familiar with) extends through summer 2014. Both tables supposedly originate with NOAA CPC. I’m not sure why the former one (that MGET uses) is not more up to date. If this is a problem, could you try
contacting the NOAA ESRL PSD folks that host the oni.data page? If you do that, please CC me.
2.
Iterate through the table looking for spans of five or more consecutive months with values >= 0.5 or <= -0.5 and classify them as El Nino or La Nina episodes respectively.
Spans of four or fewer months, and months with values between -0.5 and 0.5 are “normal” episodes. The five month threshold was based on the documentation on the ensoyears.html page.
Does that answer your question?
I have not looked at the detrend.nino34.ascii.txt file. It sounds like that is the basis from which
CPC derives ONI. In any case, MGET does not read that file.
Jason
From: Arliss Winship - NOAA Affiliate
[mailto:]
Sent: Monday, November 03, 2014 10:21 AM
To: Bryan Costa - NOAA Affiliate
Cc:
Subject: [mget-help] Re: Question about ENSO monthly binning
Yeah, I was mainly wondering whether the MGET tool uses the ONI index (and phases) from the table shown here:
http://www.cpc.ncep.noaa.gov/products/analysis_monitoring/ensostuff/ensoyears.shtml
Or if it calculates the ONI and identifies phases on the basis of the methods described on that webpage along with the monthly data served here:
The reason I ask is that the latter can produce slightly different results (e.g., phases) depending on the numerical precision of the calculations.
On Mon, Nov 3, 2014 at 10:15 AM, Bryan Costa - NOAA Affiliate <> wrote:
Hi Jason,
I'm writing because I was hoping to get more information about how MGET tools use the
ONI table to bin rasters into ENSO phases (Arliss, please feel to jump in).
The reason that I ask is we'd like to bin other oceanographic products (namely eddy probabilities and HYCOM surface currents/temp) in the same way, and would like to make sure our
methods are consistent across all our predictor variables.
Bryan Costa
--
NOAA|CCMA|Biogeography Branch
N-SCI-1, SSMC 4, 9th Floor, #9232
The contents of this message are mine and do not necessarily reflect any position of NOAA
--
NOAA National Ocean Service (NOS)
National Centers for Coastal Ocean Science (NCCOS)
Center for Coastal Monitoring and Assessment (CCMA)
Contractor, CSS-Dynamac Inc.
1305 East-West Hwy, SSMC-4, N/SCI-1, #9245
*******************************************
Disclaimer: Any views or opinions expressed in this message are those of the sender, and do not represent official views of NOAA, the United States Government or its agents, or
CSS-Dynamac Inc.
--
NOAA National Ocean Service (NOS)
National Centers for Coastal Ocean Science (NCCOS)
Center for Coastal Monitoring and Assessment (CCMA)
Contractor, CSS-Dynamac Inc.
1305 East-West Hwy, SSMC-4, N/SCI-1, #9245
Tel: +1-(301)-713-3028 ext. 200
*******************************************
Disclaimer: Any views or opinions expressed in this message are those of the sender, and do not represent official views of NOAA, the United States Government or its agents, or
CSS-Dynamac Inc.
--
NOAA|CCMA|Biogeography Branch
N-SCI-1, SSMC 4, 9th Floor, #9232
Phone: (301) 713-3028 x146
The contents of this message are mine and do not necessarily reflect any position of NOAA
--
NOAA|CCMA|Biogeography Branch
N-SCI-1, SSMC 4, 9th Floor, #9232
Phone: (301) 713-3028 x146
The contents of this message are mine and do not necessarily reflect any position of NOAA
|