Difference between revisions of "Forecast Data"

From SCECpedia
Jump to navigationJump to search
 
(29 intermediate revisions by the same user not shown)
Line 1: Line 1:
 
Links to existing SCEC Forecast Data Distribution Methods:
 
Links to existing SCEC Forecast Data Distribution Methods:
 +
 +
== ETAS Forecast Data Prototype Site ==
 +
* [https://forecast.derivacloud.org/chaise/recordset/#5/ETAS:Forecast@sort(RID) ETAS Forecast Data and Evaluations]
 +
== Related Work ==
 +
*[https://cmu-delphi.github.io/covidcast/talks/evalcast/talk.html#(1) Covid-cast Example]
 +
*[https://mybinder.org/ Binder]
 +
*[https://www.nature.com/articles/s41597-020-0486-7 Extending FAIR Data to TRUST Principles]
 +
 +
== Earthquake Prediction Projects ==
 +
*[https://www.wired.co.uk/article/ai-predicting-earthquakes AI Predictions]
 +
*[https://aeta.io/ AETA Earthquake Prediction Project]
 +
*[https://hmsc.harvard.edu/-computational-scientist-brendan-meade Brendan Meade Ml Forecasting]
 +
*[https://arxiv.org/abs/1507.03228 Meade Conference Paper]
 +
*[http://approximateinference.org/2017/accepted/LindermanEtAl2017.pdf Meade Conference Paper]
 +
 +
== Existing ETAS Data Management Systems ==
 +
*[http://opensha.usc.edu/ftp/kmilner/markdown/ucerf3-etas-results/2019_09_04-ComCatM7p1_ci38457511_ShakeMapSurfaces/index.html#comcat-data-comparisons UCERF3-TD ETAS Studies]
 +
*[https://github.com/cseptesting/ridgecrest_evaluation_bssa Github Repo with CSEP Evaluations of ETAS forecasts]
 +
*[/home/scec-06/kmilner/ucerf3/etas_sim/2019_09_04-ComCatM7p1_ci38457511_ShakeMapSurfaces/ Milner et al 2020 Ridgecrest ETAS Forecasts]
 +
*[/home/scec-06/wsavran/working/ucerf3-ridgecrest Savran et al 2020 Ridgecrest ETAS Forecast]
 +
 +
== Example Results Directory ==
 +
<pre>
 +
-bash-4.2$ du -sh 2019_09_04-ComCatM7p1_ci38457511_ShakeMapSurfaces
 +
36G 2019_09_04-ComCatM7p1_ci38457511_ShakeMapSurfaces
 +
 +
-bash-4.2$ ls -alh 2019_09_04*
 +
2019_09_04-ComCatM7p1_ci38457511_ShakeMapSurfaces:
 +
total 36G
 +
drwxr-sr-x  3 kmilner lc_pjm 4.0K Oct 31 10:10 .
 +
drwxr-sr-x 455 kmilner lc_pjm  40K Dec  7 10:20 ..
 +
drwxr-sr-x  2 kmilner lc_pjm 4.0K Oct 31 10:10 config_input_plots
 +
-rw-r--r--  1 kmilner lc_pjm  77K Oct 31 10:06 config.json
 +
-rw-r--r--  1 kmilner lc_pjm 2.5K Oct 31 10:10 etas_sim_mpj.slurm
 +
-rw-r--r--  1 kmilner lc_pjm    0 Oct 31 10:10 etas_sim_mpj.slurm.e4413437
 +
-rw-r--r--  1 kmilner lc_pjm  66M Oct 31 09:55 etas_sim_mpj.slurm.o4413437
 +
-rw-r--r--  1 kmilner lc_pjm  70M Oct 31 09:53 opensha-ucerf3-all.jar
 +
-rw-r--r--  1 kmilner lc_pjm  36G Oct 30 23:37 results_complete.bin
 +
-rw-r--r--  1 kmilner lc_pjm 162M Oct 31 09:48 results_m5_preserve_chain.bin
 +
</pre>
 +
 +
== Github repo ==
 +
Project source code and files are collected in a SCEC public github repository.
 +
*[https://github.com/SCECcode/scec-forecast-data.git Forecast Data Management Github Repo]
  
 
== Forecast Data Descriptions ==
 
== Forecast Data Descriptions ==
Line 8: Line 52:
 
== Forecast Data Listings ==
 
== Forecast Data Listings ==
 
ETAS Simulation Data:  
 
ETAS Simulation Data:  
 +
*[http://opensha.usc.edu/ftp/kmilner/markdown/ucerf3-etas-results/2019_09_04-ComCatM7p1_ci38457511_ShakeMapSurfaces/index.html#comcat-data-comparisons ShakeMap Surface Comparisons]
 
*[http://opensha.usc.edu/ftp/kmilner/markdown/ucerf3-etas-results/ OpenSHA ETAS]
 
*[http://opensha.usc.edu/ftp/kmilner/markdown/ucerf3-etas-results/ OpenSHA ETAS]
 
*[https://github.com/kevinmilner/ucerf3-etas-results/tree/master/2020-weekly-runs/2020_05_14-weekly-1986-present-full_td-kCOV1.5#table-of-contents Weekly Summary]
 
*[https://github.com/kevinmilner/ucerf3-etas-results/tree/master/2020-weekly-runs/2020_05_14-weekly-1986-present-full_td-kCOV1.5#table-of-contents Weekly Summary]
 +
*[http://opensha.usc.edu/ftp/kmilner/markdown/ucerf3-etas-results/2019_09_04-ComCatM7p1_ci38457511_ShakeMapSurfaces/index.html#comcat-data-comparisons ShakeMap Summary]
  
 
UCERF Project Data:
 
UCERF Project Data:
Line 39: Line 85:
  
 
== Related Projects ==
 
== Related Projects ==
 +
Data Discovery CI:
 
*[https://bd2k.ini.usc.edu/tools/deriva/ Deriva Project]
 
*[https://bd2k.ini.usc.edu/tools/deriva/ Deriva Project]
 +
*[https://eugeneyan.com/writing/data-discovery-platforms Research Community Data Discovery Platforms]
 +
Reproducibility Projects:
 
*[https://wholetale.org/ WholeTale Project]
 
*[https://wholetale.org/ WholeTale Project]
−
*[https://www.iris.edu/hq/initiatives/rapid_response Geohazard Projects]
+
Trustworthy Computing Projects:
−
*[https://www.iris.edu/hq/initiatives/rapid_response Mid-scale Research Infrastructure - 1]
+
*[https://sciencenode.org/feature/Trustworthy%20forecasts.php Trustworthy Forecast Project]
−
*[https://sciencenode.org/feature/Trustworthy forecasts.php?clicked=title Trustworthy Forecast Project]
+
Software Institute
 
*[https://molssi.org/ Software Institute]
 
*[https://molssi.org/ Software Institute]
−
*[https://www.nsf.gov/geo/geo-ci/index.jsp NSF Geoscience opportunities]
+
Rapid Hazard Response Projects
−
*[https://www.tulane.edu/~sanelson/Natural_Disasters/introduction.htm Geohazard Overview]
+
*[https://www.iris.edu/hq/initiatives/rapid_response Rapid Response Projects]
−
*[https://eugeneyan.com/writing/data-discovery-platforms Data Discovery Platforms]
+
Time Series Forecasting Service
 
*[https://medium.com/oreillymedia/the-future-of-time-series-forecasting-bd83c2aca9a8 Time Series Forecasting Services]
 
*[https://medium.com/oreillymedia/the-future-of-time-series-forecasting-bd83c2aca9a8 Time Series Forecasting Services]
 +
Controlled environment for exchanging and comparing optimization algorithms.
 +
*[https://core.isrd.isi.edu/id/1/W1QR@2VC-3Z9S-4D36 Coredev Project Site]
 +
 +
== Seismic Hazard User Base Estimates ==
 +
*[https://www.iris.edu/hq/initiatives/rapid_response Geohazard Projects]
 +
*[https://www.tulane.edu/~sanelson/Natural_Disasters/introduction.htm Geohazard Research and Users Community Overview]
  
−
==Forecast User Base Estimates ==
 
 
*We have these estimates of the user base for products that come out of CISM and CSEP. And how those numbers are documented. UCERF3-ETAS: 5-10 people currently look at raw simulation results. If we create a full operational system, it would potentially have broad user base, but it's currently in a research mode and only shared with interested and sophisticated users.
 
*We have these estimates of the user base for products that come out of CISM and CSEP. And how those numbers are documented. UCERF3-ETAS: 5-10 people currently look at raw simulation results. If we create a full operational system, it would potentially have broad user base, but it's currently in a research mode and only shared with interested and sophisticated users.
 
*UCERF3+ Inversions: The inversion results will be looked at by the core WGCEP team and shared with the scientific review panel, so probably 5-10 diving in and looking at detailed results, ~30 looking at summary plots. The New Zealand team would likely be interested in these capabilities as well, which could double the user base.RSQSim: Likely always a small group, <10 looking at RSQSim results and downstream products (ground motion simulations)
 
*UCERF3+ Inversions: The inversion results will be looked at by the core WGCEP team and shared with the scientific review panel, so probably 5-10 diving in and looking at detailed results, ~30 looking at summary plots. The New Zealand team would likely be interested in these capabilities as well, which could double the user base.RSQSim: Likely always a small group, <10 looking at RSQSim results and downstream products (ground motion simulations)
 
*With the RISE experiment gearing up and many new modelers looking to get their models tested, and assuming that UCERF3/4 goes under prospective testing I expect between 20-30 people will be interested in looking at individual forecast and evaluation results. An additional ~10 would be interested in looking at aggregate results, such as the cumulative tests (ie., types of results that could be a publication figure), but I don’t have a good handle on those numbers. Like with the UCERF3 results, the potential user base would probably increase depending on the availability of the data products and simulations results.
 
*With the RISE experiment gearing up and many new modelers looking to get their models tested, and assuming that UCERF3/4 goes under prospective testing I expect between 20-30 people will be interested in looking at individual forecast and evaluation results. An additional ~10 would be interested in looking at aggregate results, such as the cumulative tests (ie., types of results that could be a publication figure), but I don’t have a good handle on those numbers. Like with the UCERF3 results, the potential user base would probably increase depending on the availability of the data products and simulations results.
  
−
== NSF MRI-1 Solicitation Site ==
+
== NSF Solicitations Site ==
−
*[https://www.nsf.gov/pubs/2021/nsf21505/nsf21505.htm Solicitation]
+
*[https://www.nsf.gov/pubs/2021/nsf21505/nsf21505.htm Mid-scale Research Infrastructure - 1 Solicitation]
 +
*[https://www.nsf.gov/pubs/2021/nsf21537/nsf21537.htm Mid-scale Research Infrastructure - 2 Solicitation]
 +
*[https://www.nsf.gov/geo/geo-ci/index.jsp NSF Geoscience opportunities]
 +
*[https://www.nsf.gov/cise/harnessingdata/ HarnessingDataRevolution]
 +
*[https://www.nsf.gov/pubs/2020/nsf20091/nsf20091.jsp NSF CISE DCL Big Ideas]
  
 
== Related Entries ==
 
== Related Entries ==
 
*[[CME_Project]]
 
*[[CME_Project]]

Latest revision as of 00:32, 26 June 2021

Links to existing SCEC Forecast Data Distribution Methods:

ETAS Forecast Data Prototype Site

Related Work

Earthquake Prediction Projects

Existing ETAS Data Management Systems

Example Results Directory

-bash-4.2$ du -sh 2019_09_04-ComCatM7p1_ci38457511_ShakeMapSurfaces 
36G	2019_09_04-ComCatM7p1_ci38457511_ShakeMapSurfaces

-bash-4.2$ ls -alh 2019_09_04*
2019_09_04-ComCatM7p1_ci38457511_ShakeMapSurfaces:
total 36G
drwxr-sr-x   3 kmilner lc_pjm 4.0K Oct 31 10:10 .
drwxr-sr-x 455 kmilner lc_pjm  40K Dec  7 10:20 ..
drwxr-sr-x   2 kmilner lc_pjm 4.0K Oct 31 10:10 config_input_plots
-rw-r--r--   1 kmilner lc_pjm  77K Oct 31 10:06 config.json
-rw-r--r--   1 kmilner lc_pjm 2.5K Oct 31 10:10 etas_sim_mpj.slurm
-rw-r--r--   1 kmilner lc_pjm    0 Oct 31 10:10 etas_sim_mpj.slurm.e4413437
-rw-r--r--   1 kmilner lc_pjm  66M Oct 31 09:55 etas_sim_mpj.slurm.o4413437
-rw-r--r--   1 kmilner lc_pjm  70M Oct 31 09:53 opensha-ucerf3-all.jar
-rw-r--r--   1 kmilner lc_pjm  36G Oct 30 23:37 results_complete.bin
-rw-r--r--   1 kmilner lc_pjm 162M Oct 31 09:48 results_m5_preserve_chain.bin

Github repo

Project source code and files are collected in a SCEC public github repository.

Forecast Data Descriptions

Forecast Data Listings

ETAS Simulation Data:

UCERF Project Data:

CSEP Project Data:

Ground Motion Simulation Project Data:

Event Response Data:

Broadband Platform Data:

Data Sharing Methods

Related Projects

Data Discovery CI:

Reproducibility Projects:

Trustworthy Computing Projects:

Software Institute

Rapid Hazard Response Projects

Time Series Forecasting Service

Controlled environment for exchanging and comparing optimization algorithms.

Seismic Hazard User Base Estimates

  • We have these estimates of the user base for products that come out of CISM and CSEP. And how those numbers are documented. UCERF3-ETAS: 5-10 people currently look at raw simulation results. If we create a full operational system, it would potentially have broad user base, but it's currently in a research mode and only shared with interested and sophisticated users.
  • UCERF3+ Inversions: The inversion results will be looked at by the core WGCEP team and shared with the scientific review panel, so probably 5-10 diving in and looking at detailed results, ~30 looking at summary plots. The New Zealand team would likely be interested in these capabilities as well, which could double the user base.RSQSim: Likely always a small group, <10 looking at RSQSim results and downstream products (ground motion simulations)
  • With the RISE experiment gearing up and many new modelers looking to get their models tested, and assuming that UCERF3/4 goes under prospective testing I expect between 20-30 people will be interested in looking at individual forecast and evaluation results. An additional ~10 would be interested in looking at aggregate results, such as the cumulative tests (ie., types of results that could be a publication figure), but I don’t have a good handle on those numbers. Like with the UCERF3 results, the potential user base would probably increase depending on the availability of the data products and simulations results.

NSF Solicitations Site

Related Entries

Retrieved from "https://strike.scec.org/scecwiki/index.php?title=Forecast_Data&oldid=25678"