30
Argo Status & issues for ADMT Argo TC ADMT#13 November 2012, Hyderabad

Argo Status & issues for ADMT Argo TC

Embed Size (px)

DESCRIPTION

Argo Status & issues for ADMT Argo TC. ADMT#13 November 2012, Hyderabad. Argo has been sustaining a 3000 floats array for 4 years and is starting improving it …. 12 nations maintain the global array and 20 more fill regional gaps 18 countries active in 2012. - PowerPoint PPT Presentation

Citation preview

Page 1: Argo Status & issues for ADMT  Argo TC

Argo Status& issues for ADMT

Argo TC

ADMT#13November 2012, Hyderabad

Page 2: Argo Status & issues for ADMT  Argo TC

212 nations maintain the global array and 20 more fill regional gaps18 countries active in 2012

Argo has been sustaining a 3000 floats array for 4 years and is starting improving it…

Page 3: Argo Status & issues for ADMT  Argo TC

3

However more international cooperation is required!

50+% maintained by the USA … Growing involvement of Europe is crucial

Argo should be promoted further in Russia, Mexico, Indonesia, Turkey, Taiwan …All these countries have already cooperated.

New commitments for 2013: Brazil (2 prog.), Iran, Taiwan

Sri Lanka (redeployed a Chinese float)

Page 4: Argo Status & issues for ADMT  Argo TC

After a drop around 2010, Argo was back to its target in 2011we can be very optimistic for 2012

4

Time to expand our “Core Argo” to the “Global” (AST#13)What would be the number for a “Global Argo” 3° x 3° ? Þ New global index ? Regional indexes ? Deep Argo indexes ?Þ Refined “Customer” requirements (e.g. Mercator)Þ See ASW#4 report

Page 5: Argo Status & issues for ADMT  Argo TC

5If we consider a global coverage and remove those floats,

Core Argo: new definition …

Page 6: Argo Status & issues for ADMT  Argo TC

6If we consider profiles pressure >1750 dbar

Core Argo: new definition …

Page 7: Argo Status & issues for ADMT  Argo TC

775 % of the array is meeting requirements (beyond spatial distribution)

We obtain a new « Argo core » value (target remains to be defined)Challenge is still forward…

Page 8: Argo Status & issues for ADMT  Argo TC

But there is still the 2009 deployments deficit to addressIt may require a couple more years

8

Challenging Areas: S.O.

North West I.O.

Large parts of the P.O.

Page 9: Argo Status & issues for ADMT  Argo TC

Challenge in logistics is still impressing ….

Page 10: Argo Status & issues for ADMT  Argo TC

Challenge in logistics is still impressing ….

Page 11: Argo Status & issues for ADMT  Argo TC

11A large part of the Pacific Ocean will need to be covered within a couple of yearsas well as the W Atlantic and NE/W Indian and SO

How optimize the network coverage now ?

Page 12: Argo Status & issues for ADMT  Argo TC

12

… and tomorrow ?

Page 13: Argo Status & issues for ADMT  Argo TC

13

How the Argo teams are going to fill the gaps ? Deployment opportunities ?

Dedicated opportunities and funding required

see e.g. Karahoa, Lady Amber

& partnerships with Industry, Sailing, Foundations , NGOs (VSF), etc.

New JCOMMOPS Ship Coordinator (official Nov. 19th)

SOT (VOS, SOOP, GOSHIP) TC

Page 14: Argo Status & issues for ADMT  Argo TC

14

A good diversity of float models and manufacturers ( 7 commercial)2/3 of the market is handled by Teledyne

Improving reliabilityWarning for low cost instruments …

Page 15: Argo Status & issues for ADMT  Argo TC

15

Iridium used on ¼ of 2011 deployments, probably 40% in 2012

Diversity is important …

Page 16: Argo Status & issues for ADMT  Argo TC

Argo Data Stream: All profiles1 000 000th profile early Nov. 20121 profile every 5-6 minute …

16

The difference GDACS/GTS profiles is rather stable (around 20% every year). If a number of profiles cannot be inserted on GTS because floats are greylisted, or floats data are released lately , we can still optimize GTS distribution volume by 10%.

An effort on the processing of first profiles in time (with pre registration done, decoders tested, etc) will help.

Page 17: Argo Status & issues for ADMT  Argo TC

1000000th profile float

17

Many ways to count profiles and counting may change everyday

AIC countdown to AST: ordering profiles from the registered floats by obs. date

Homage to ADMT13 host India.

By the way, a few dots on land …

Page 18: Argo Status & issues for ADMT  Argo TC

DM Status (see specific agenda item and AIC report for details)

18

DM/DM Eligible03/2010: 76% 03/2011: 83%03/2012: 82%11/2012: 80%

Only 6% are not under control

Page 19: Argo Status & issues for ADMT  Argo TC

19

Remark: Argo/Argo eq. is a small and stable fraction of the array (10%). To keep in mind for RT/DM data processing.

Page 20: Argo Status & issues for ADMT  Argo TC

• Float pending at AIC (74): a bit highFloats (iridium tracking/decoding issues)

• Float not yet registered at AIC (21): reasonableBSH, KORDI, Coriolis (DEEP_ARVOR), WHOI, PMEL, UK, UW

• Float not yet on GTS (186): too high INCOIS, BSH and NAVO

• Float not yet at GDACS (51): usualMany BSH floats (Coriolis)

RT Issues

Page 21: Argo Status & issues for ADMT  Argo TC

Why important to study ?

• Meet operational requirements (24h)

• Optimize and harmonize data flow and DAC practices

• Detect errors, issues, failures, blocking points in data processing chains

Þ See ISDM agenda item for GTS

Þ See AIC dedicated report for DACs/GDACs Effort to be continued, with DAC feedbackstats to be refined

Delays

Page 22: Argo Status & issues for ADMT  Argo TC

Delays at GDACs: Where are we now? Not too bad … but still room for progress

Page 23: Argo Status & issues for ADMT  Argo TC

23

Delays at GDACs over 2012

Page 24: Argo Status & issues for ADMT  Argo TC

24

Delays at GDACs over 2012

A large number of files is resubmitted in delayed-mode

Processing/Decoding new floats, or eq. floats impact delays

(see diagonals)

Green dots show real-time profiles

General and impressive improvement over 2nd half of 2012? GDAC improvement ?GDAC should not update GDAC_CREATION_DATE in index when DM file is submitted (?)

Strange gap around Feb. 11th (CLS ?)See all DACs details in report and feedback if needed

Page 25: Argo Status & issues for ADMT  Argo TC

• The Delays observed at GDACs have improved substantially in the second half of 2012 through a combined effort at DACs and GDAC level.

• There is some room for progress for a few DACs

• There are too much late submissions of profiles. If median value (29h) is reasonable, only 36% of observations are distributed to users within 24h. This impacts in particular large areas of the oceans.

• Cooperation between national DACs and redundancy are encouraged• Cooperation (processing chains, decoders) between DACs is encouraged (see CSIRO and INCOIS, KORDI)

• Float STANDARD_FORMAT_ID to circulate before deployment, as well as data samples for testing

• A number of processing steps are launched at different time, with rough offsets, which introduce small delays. They should rather be chained.

• It would be good to set up a RT raw data delivery from Argos (see Coriolis recent practice) coupled to a RT processing chain.

• Special floats (NOVA, NEMO, etc) requesting preprocessing steps should be monitored in detail and optimized if needed, or integrated to existing chains.

• AIC capacity to generate metafiles from notification step (and FTP to GDACs if needed) or use metafiles for notification will help (from PIs or GDACs). PI notification to DAC and AIC should happen at same time. Automatic crawling/upload systems would be best. .

Report conclusions

Page 26: Argo Status & issues for ADMT  Argo TC

• The number of countries involved in Argo is good but not sufficient.Argo would be impacted immediately if USA has difficultiesThere is still room for a few more big players.

• 2009 deficit is still not solved.Resources required for ship time, and organization?Need to plan big deployment batches (efficient) and flexible cruises

• We need better indexes to track Argo network(s) progress

• Delays in data availability at the GDACs have been partially solved.AIC survey of DACs practices should be continued.

• Float lifetime keeps improving. Beware of low cost instrumentation!

• New generation of telecoms seems ready on all float models. Groups experiencing. Needs to be more operational and centralized.

Conclusion

Page 27: Argo Status & issues for ADMT  Argo TC

• Data flow can be improved from deployment to final formatting– Too many discontinuous steps– Too many human interventions

• Better delays, more data shared in RT

• deployment planning information management (between PIs, AIC, DACs) should flow machine to machine (interoperability)

• Sharing same vocabulary on metadata everywhere, and introducing the STANDARD_FORMAT_ID as soon as possible in the data

flow is crucial(see Esmee agenda item)

Why not introduce our WMO ID at manufacturer level (PMEL proposal) ?

Conclusion

Page 28: Argo Status & issues for ADMT  Argo TC

• Is our FTP/netCDF based data exchange format optimized ?

• Short/medium termncISO ? See agenda item on (ACDD and CF conventions)Will we do all changes required? (pressure from NOAA)US NODC to make another Argo netCDF ?

• Should we start preparing the next generation today after a decade of good services?

• Is our GDAC “mirroring” optimal ?

• What is the size of our user community ? 500 users/day ?• How to expand it outside the Argo community?

These issues should be debated openly and long run anticipated

Questions

Page 29: Argo Status & issues for ADMT  Argo TC

• Workshop on data access, data use, data display from a technical perspective targeting interoperability, monitoring , outreach, education, etc.

• Many tools, software available (desktop, web, GIS, etc)

• Private and external actors coming to fill up gaps

• Some limitation with growing data set have to be anticipated

• And new tools prepared for the future

– Aside an ADMT meeting e.g.– Or separately (JCOMMOPS offers)

Suggestion

Page 30: Argo Status & issues for ADMT  Argo TC

30

[email protected]@jcommops.org