incubator-cvs mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From Apache Wiki <wikidi...@apache.org>
Subject [Incubator Wiki] Trivial Update of "ClimateProposal" by AndrewHart
Date Sat, 26 Jan 2013 18:33:08 GMT
Dear Wiki user,

You have subscribed to a wiki page or wiki category on "Incubator Wiki" for change notification.

The "ClimateProposal" page has been changed by AndrewHart:
http://wiki.apache.org/incubator/ClimateProposal?action=diff&rev1=17&rev2=18

Comment:
clarify reference in as specific to RCMED in P1 of abstract (...was RCMES) 

  === Abstract ===
  The Apache Open Climate Workbench proposal desires to contribute an existing community of
software related to the analysis and evaluation of climate models, and related to the use
of remote sensing data in that process. 
  
- Specifically, we will bring a fundamental software toolkit for analysis and evaluation of
climate model output against remote sensing data. The toolkit is called the [[http://rcmes.jpl.nasa.gov|Regional
Climate Model Evaluation System (RCMES)]]. RCMES provides two fundamental components for the
easy, intuitive comparison of climate model output against remote sensing data. The first
component called RCMED (for "Regional Climate Model Evaluation Database") is a scalable cloud
database that decimates remote sensing data and renalysis data related to climate using Apache
OODT extractors, Apache Tika, etc. These transformations make traditionally heterogeneous
upstream remote sensing data and climate model output homogeneous and unify them into a data
point model of the form (lat, lng, time, value, height) on a per parameter basis. Latitude
(lat) and Longitude (lng) are in WGS84 format, but can be reformatted on the fly. time is
in ISO 8601 format, a string sortable format independent of underlying store. value carries
with it units, related to interpretation and height allows for different values for different
atmospheric vertical levels. All of RCMES is built on Apache OODT, Apache Sqoop/Apache Hadoop
and Apache Hive, along with hooks to PostGIS and MySQL (traditional relational databases).
The second component of the system, RCMET (for "Regional Climate Model Evaluation Toolkit")
provides facilities for connecting to RCMED, dynamically obtaining remote sensing data for
a space/time region of interest, grabbing associated model output (that the user brings, or
from the Earth System Grid Federation) of the same form, and then regridding the remote sensing
data to be on the model output grid, or the model output to be on the remote sensing data
grid. The regridded data spatially is then temporally regridded using techniques including
seasonal cycle compositing (e.g., all summer months, all Januaries, etc.), or by daily, monthly,
etc. The uniform model output and remote sensing data are then analyzed using pluggable metrics,
e.g., Probability Distribution Functions (PDFs), Root Mean Squared Error (RMSE), Bias, and
other (possibly user-defined) techniques, computing an analyzed comparison or evaluation.
This evaluation is then visualized by plugging in to the NCAR NCL library for producing static
plots (histograms, time series, etc.)
+ Specifically, we will bring a fundamental software toolkit for analysis and evaluation of
climate model output against remote sensing data. The toolkit is called the [[http://rcmes.jpl.nasa.gov|Regional
Climate Model Evaluation System (RCMES)]]. RCMES provides two fundamental components for the
easy, intuitive comparison of climate model output against remote sensing data. The first
component called RCMED (for "Regional Climate Model Evaluation Database") is a scalable cloud
database that decimates remote sensing data and renalysis data related to climate using Apache
OODT extractors, Apache Tika, etc. These transformations make traditionally heterogeneous
upstream remote sensing data and climate model output homogeneous and unify them into a data
point model of the form (lat, lng, time, value, height) on a per parameter basis. Latitude
(lat) and Longitude (lng) are in WGS84 format, but can be reformatted on the fly. time is
in ISO 8601 format, a string sortable format independent of underlying store. value carries
with it units, related to interpretation and height allows for different values for different
atmospheric vertical levels. All of RCMED is built on Apache OODT, Apache Sqoop/Apache Hadoop
and Apache Hive, along with hooks to PostGIS and MySQL (traditional relational databases).
The second component of the system, RCMET (for "Regional Climate Model Evaluation Toolkit")
provides facilities for connecting to RCMED, dynamically obtaining remote sensing data for
a space/time region of interest, grabbing associated model output (that the user brings, or
from the Earth System Grid Federation) of the same form, and then regridding the remote sensing
data to be on the model output grid, or the model output to be on the remote sensing data
grid. The regridded data spatially is then temporally regridded using techniques including
seasonal cycle compositing (e.g., all summer months, all Januaries, etc.), or by daily, monthly,
etc. The uniform model output and remote sensing data are then analyzed using pluggable metrics,
e.g., Probability Distribution Functions (PDFs), Root Mean Squared Error (RMSE), Bias, and
other (possibly user-defined) techniques, computing an analyzed comparison or evaluation.
This evaluation is then visualized by plugging in to the NCAR NCL library for producing static
plots (histograms, time series, etc.)
  
  We also have performed a great deal of work in packaging RCMES to make the system easy to
deploy. We have working Virtual Machines (VMWare VMX and Virtual Box OVA compatible formats)
and we also have an installer built on Python Buildout (http://buildout.org/) called "Easy
RCMET" for dynamically constructing the RCMET toolkit.
  
@@ -36, +36 @@

  === Initial Goals ===
  The initial goals of the proposed project are:
  
-  * Stand up a sustaining Apache-based community around the JPL RCMES codebase.
+  * Stand up a sustaining Apache-based community around the JPL RCMES and CDX codebases.
+  * Active relationships and possible cooperation with related projects and communities.
+  * more.
+ 
-  * Active relationships and possible cooperation with related projects and communities,
including end user and scientific communities, CORDEX, NARCCAP, US NCA, IPCC, ESG, etc.
-  * Active relationships and possible cooperation with existing Apache communities, e.g.,
OODT, Hadoop/HIVE, Sqoop, Tika, SIS, etc.
-  * Initial Apache release.
-  * Leverage Apache Open Climate Toolkit in climate activities at NASA, in the international
community as mentioned above, and beyond.
-  * Vetting all software licenses and making sure IP is clear (software grant from JPL forthcoming).
-  
  == Current Status ==
  === Meritocracy ===
- 30% of the proposed initial committers are familiar with the meritocracy principles of Apache.
As state above this includes 5 ASF members. Of the mentorship list, we have included Chris
Douglas, a PMC member from Hadoop and ASF member to help guide the community. Chris M. and
Chris D. have guided a number of projects through the Incubator over the years. The 70% of
proposed initial committers that aren't as familiar with Apache have a broad range of experience
in other open source projects, and have a deep respect and affinity for the foundation and
the work that gets done here.
+ Most of the proposed initial committers are familiar with the meritocracy principles of
Apache...say more.
  === Community ===
  There is an existing, established community of developers and users of CLIMATE -- say more.
  

---------------------------------------------------------------------
To unsubscribe, e-mail: cvs-unsubscribe@incubator.apache.org
For additional commands, e-mail: cvs-help@incubator.apache.org


Mime
View raw message