| A | B | C | D | E | F | G | H | I | J | K | L | M | N | O | P | Q | R | S | T | U | V | W | X | Y | Z | AA | AB | ||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
1 | To Characterize Different Types of Earth | Science Data Analytics | What information will | help us do that? | What is informatiom | asking for? | |||||||||||||||||||||||
2 | Use Case Information: | Definition/Description | Ethan's Use Case 1 | Robert's Use Case 1 | Tiffany's Use Case 1 | Tiffany's Use Case 2 | Steve's Use Case 1 | Steve's Use Case 2 | Laura's Use Case | Chung-Lin's Use Case 1 | Chung-Lin's Use Case 2 | Comments | |||||||||||||||||
3 | |||||||||||||||||||||||||||||
4 | Use Case Title | MERRA Analytics Services: Climate Analytics-as-a-Service | MUSTANG QA: Ability to detect seismic instrumentation problems | Inter-calibrations among datasets | Inter-comparisons between multiple model or data products | Sampling Total Precipitable Water Vapor using AIRS and MERRA | Using Earth Observations to Understand and Predict Infectious Diseases | CREATE-IP - Collaborative REAnalysis Technical Environment - Intercomparison Project | The GSSTF Project (MEaSUREs-2006) | Science- and Event-based Advanced Data Service Framework at GES DISC | |||||||||||||||||||
5 | Author/Company/Email | author of the use case | Ethan McMahon, US EPA, mcmahon.ethan@epa.gov | Robert Casey, IRIS rob@iris.washington.edu | Tiffany Mathews, SSAI (@ NASA ASDC) tiffany.j.mathews@nasa.gov | Tiffany Mathews, SSAI (@ NASA ASDC) tiffany.j.mathews@nasa.gov | Steve Kempler, NASA/GSFC, Steven.J.Kempler@nasa.gov | Steve Kempler, NASA/GSFC, Steven.J.Kempler@nasa.gov | Gerald L. Potter, Laura Carriere laura.carriere@nasa.gov | Chung-Lin Shie, UMBC/JCET, NASA/GSFC Chung-Lin.Shie-1@nasa.gov | Chung-Lin Shie, UMBC/JCET, NASA/GSFC Chung-Lin.Shie-1@nasa.gov | ||||||||||||||||||
6 | Actors/Stakeholders/Project URL and their roles and responsibilities | Who is interested in the project (perhaps providing guidance) and who will benefit from the project (recipients, users, beneficiaries, etc.) | Commercial parties that could use climate model data, such as air conditioning manufacturers, water vendors, farmers and investors. These parties could project the demand for their products and services. NASA Goddard is offering the data to whomever may want to use it. | IRIS Instrumentation Services; IRIS Global Seismic Network (GSN); IRIS Quality Assurance Working Group. Features are currently deployed at http://service.iris.edu/mustang in the form of web services. | Committee on Earth Observation Satellites (CEOS) Working Group on Calibration and Validation (WGCV); World Meteorological Organization (WMO); Global Space based Inter-Calibration System (GSICS) community; Instrument and Science Teams | Science and Instrument Teams Researchers (Academia) | GMAO; Reseacrh scientists | Public Health Decision Makers | Five major reanalysis projects - NASA's MERRA; ECMWF's ERA-Interim; NOAA/NCEP's CFSR; NOAA/ESRL's 20CR; JMA's JRA-25 and JRA-55. | Data producers/providers; Research scientists; Community users | Data producers/providers; Research scientists; Community users | ||||||||||||||||||
7 | Use Case Goals | What is the goal of the analytics?: 1. To calibrate dataset? 2. To generate a new data product? 3. To perform course data reduction 4. To tease out science information from the data? 5. To validate data (quality)? 6. To inter-compare data? 7. To forecast/predict? 8. To derive conclusions? 9. To derive analytics tools? Other? | MERRA’s goal is a climate‐quality analysis that places NASA’s EOS observations into a climate context. Provide a library of commonly used spatiotemporal operations (canonical ops) that can be composed to enable higher-order analyses (#9) | Automatically gather various metrics on incoming and archival seismic data to assess and improve the quality of the stored seismic data as well as detect adverse conditions with the seismic instrumentation in the field. (#5 and/or #6??; and #7??) | To be able to quickly compare co-located measurements with matched viewing geometries from different sensors on separate spacecraft to improve target instrument calibration and remove temporal calibration trends. All which re-using existing science algorithms and information technology software to increase the interconnectedness of existing distributed information systems and the quality of datasets. (#1, #6, #5) | To be able to obtain high-resolution inter-comparisons between two or more model or data products in order to get more accurate measurements than what can be received from observing the global monthly means such as enabling one to make a point by point comparisons of specific locations. (#6) | AIRS and MERRA data inter-comparison (#5, #6) | Identify meteorological parameters associated with disease outbreaks; Disease forecast using meteorological parameters (#7, #8) | Reanalysis scientists are interested in reproducing the success of CMIP5 by studying reanalysis differences and similarities to improve reanalysis techniques. Reanalysis data also allows interdisciplinary scientist to compare their datasets with 30 or more years of gridded climate data. (#6) | Produce long-term (21.5 yrs) global air-sea surface turbulent fluxes gridded datasets ("hybrid" and L4) using the SSM/I multple-sensor (L2 or L3) retrievals and the NCEP reananlysis gridded prodcuts (L3 or L4). (#2) | Efficiently provide the users with a sophisticated integrated (i.e., knowledge-based) data package ("bundle") via user-friendly selecting a system-predefined science- or event-based topic (e.g., hurricane, volcano, etc.) among the currently in-developing knowledge database of the framework. | ||||||||||||||||||
8 | Use Case Description | This yields more details regarding how data analytics is utilized. Examples: 1. Datasets used, when applicable 2. Detailed analysis tasks - e.g., Combining products; Reducing data to just what is needed (subsetting); etc. 3. Results - The resulting product of the analytics. e.g., New product; New finding being sort; A decision based on some prediction | MODERN-ERA RETROSPECTIVE ANALYSIS FOR RESEARCH AND APPLICATIONS is a MERRA is a historical reprocessing of climate data. It Enables Climate Analytics-as-a-Service by combining iRODS data management, Cloudera MapReduce, and the Climate Data Services API to serve MERRA reanalysis products. | Examine one or more QA measurements to form high-level inferences as to data quality and instrument state of health. Certain metrics may highlight a number of different conditions so making use of multiple metrics tends to narrow the field as to diagnoses. Time trending and cumulative metric representations can point to other issues that may be otherwise hard to detect. | Instrument Callibration, offering a solution to instrument teams responsible for calibration and validation of target instrument data. Specifically, best inter-calibration practices for GEO and LEO sensors. | Instead of only being provided the global monthly means (averages), if users had access to two more broken down data of two or more model or data products and high-resolution inter-calibration tools, users would be better enabled to make a point by point comparisons of specific locations and better identify phenomena that contribute to certain patterns. | MERRA data is being used to cross validate with AIRS Total water vapor. Being of different characteristics, datasets need to be manipulated to enable inter-comparison. The result of the analytics will be datasets that can be more homogeneously compared. Differences can be further analyzed to assess deficiencies in both the observations and the reanalyses. | Meteorological data, TRMM precipitation and GLDAS near surface temperature and near surface specific humidity is modeled using a variety of methods to ultimately predict influenza outbreak. Uses the Disease Surveillance database for inter-comparisons | Climate reanalysis. Data sets - NASA's MERRA, ECMWF's ERA-Interim, NOAA/NCEP's CFSR, NOAA/ESRL's 20CR, and JMA's JRA-25 and JRA-55. New datasets to be added as they become available. Resulting Product - Improved climate reanalysis code. | Gobal Water and Energy Cycle. Input (applied) datasets: a) SSM/I multi-sensor (F08, F10, F11, F13, F14 and F15) brightness temperature (TB) and total precipitable water (W), surface air humidity at 10 m (Q), surface wind speed (U); b) NCEP reanalysis of sea surface/skin temperature (SST/SKT), air temperature at 2m (Tair), and sea level pressure (SLP); c) Cross-Calibrated Multi-Platform (CCMP) ocean surface wind vectors (mainly based on SSM/I multi-sensors). Output (resultant) datasets: Combined (multi sensors) and "hybrid" (satellite and model reanalysis) products consisting of new physical parameters, i.e., the air-sea turbulent fluxes: latent heat flux (LHF), sensible heat flux (SHF) and wind stress (WST). | A prototype portal related to a specific topic of Hurricane Sandy (Oct 22-31 2012) that has been developed (currently as a "recipe") at GES DISC is used here to describe the “bundle” data service. As Hurricane Sandy being selected as a user targeted case, a system-prearranged table consisting of related data variables (i.e., precipitation, winds, sea surface temperature, sea level pressure, air temperature, relative humidity, aerosols, soil moisture and surface runoff, trace gases) linked to the respective data products with fine temporal and spatial resolution of various in-house sources is provided. All the listed data variables that should be readily applied by users for studies such as hurricane track, hurricane intensity, disaster analysis, and its impacts etc. can then be readily downloaded through the data search engine, Mirador. The powerful visualization tools Giovanni (online; built in-house) is accessible for users to acquire quick and informative views of their interested data variables/products of Level 3 (gridded) in the Giovanni database. For Level 2 data (swath) and certain Giovanni-unavailable Level 3 data, the system provides a link to data recipes that give a step-by-step how-to guide to read or visualize the data using offline tools, such as Panoply, GrADS, or IDL, etc. | ||||||||||||||||||
9 | Current technical issues/requirements to take into account that may impact needed data analytics | Data Source (distributed/centralized) | Distributed - Bringing data close to the places where it is processed will speed analyses and reduce input/output concerns. | Data for analysis is centralized: an archival system with web service interface access | Data for analysis is centralized: an archival system with web service interface access | Data is within the same product or model | Data is co-located in the same facility | Distributed, but brought together | Distributed - 2D and 3D global gridded model output, multiple variables located at their respective data centers. | Involve (input) multiple datasets of multiple variables from multiple data source (e.g., multi satellites/sensors and model renanlysis) | Involve (input) multiple datasets of multiple variables from multiple data source (e.g., multi satellites/sensors and model renanlysis) | ||||||||||||||||||
10 | Volume (size) | 310 TB in total, growing 50 TB per year, accelerating | Depending on the size of the volume (to break down a global monthly means to a more specifiv point in time) the volume could become overwhelming if the appropriate tools are unavailable to identify specific times and places. | N/A | Sorage capacity problem. HOW BIG A PROBLEM | Currently 0.5 TB. Growth to 500 TB to 1 PB expected. | Currently ~1.0 TB. May grow if more funding granted. The current data has covered from July 1987-Dec 2008, i.e., 21.5 years. Assume that the PI (data provider Chung-Lin Shie) has successful won a 3-yr grant (say 2016-2018), and thus would be able to resume and extend the data produciton for another 10-yr (Jan 2009-Dec 2018). Then, a 0.5 TB data volume would be grown by Dec 2018. | The "bundle" data service here is not about to focus mainly on the data volume growth (e.g., a constantly increased data volume), but rather to provided users a knowledge-based service to efficiently (quicker and more accurately) acquire (extract) the targeted data paramenters/datasets that the users are interested with a considerably reduced volume size out of the various and originally massive data products (of big volume size). | |||||||||||||||||||||
11 | Velocity (e.g. real time) | No, because the data inputs are from the past. | Metrics gathered at rate of the data access and metrics pipeline. About 800 GB of data scanned per day. Real time data accumulates about 100 GB per day, contributing to steady source data growth. Some data arrives as an entire package, many GB to TB at a time, for any point in time, resulting in bursty growth of the archive. | NA | N/A | N/A | Data is generated on supercomputers. Most of the data is currently available or will be made available as soon as the model runs have completed. | N/A | N/A | ||||||||||||||||||||
12 | Variety - Bringing distributed heterogeneous data together (data formats, data types) | Seismic data being analyzed, along with mass position and pressure readings are currently returned in uniform fashion. Variety comes in the nature of the instrumentation, their response curves, whether triggered or not, and their colocation to other seismic stations. Native data format is SEED, an FDSN seismic standard for more than 25 years. Newly introduced are markups and representations of this data amenable to web services, among these a new XML metadata standard. Text tabular representations of data are also offered via web services. | Data from Low Earth Orbiting (LEO) and Geosynchronous Orbiting (GEO) satellite data looking at the same areas. | Heterogeneous (as it comes from the same product but is more broken down) | Heterogeneous datasets are made to be more homogeneous so they can be inter-compared: Grid, resolutiion, etc. | Heterogeneous datasets are ingested into models that can accommodate the dataset | Native formats from centers differ (netcdf, grib, variable names, etc). CREATE-IP will prepare the data in ESGF's standard format to facilitate comparison. Netcdf. (Some data is originally grib and hdf.) | Using heterologeneous data (mainly in binary format) and creating a new data prodcut (physical quantities) of HDF-EOS5 format. | "Bundle" together the diverse (yet knowledge-based) data paramenters of interests from various data products. | ||||||||||||||||||||
13 | Veracity (Robustness Issues) / Data Quality | Veracity is one of the key issues being addressed by MUSTANG, where the issues with robustness of the dataset are many. Bad data encoding, gappy data, missing data, incorrect metadata entries, bad time clock values, and overlapping datasets are all common problems encountered and must be accounted for and corrected. Real time data feeds generally have small flaws that are later corrected by a followup, post-processed data set. The Quality Assurance Working Group for IRIS is charged with answering the question of what constitutes quality seismic data. MUSTANG is meant to generate figures that can help to detect a number of anomalies. If anomalies are few or not detected, then this could be one condition of flagging quality data. The data must cleanly and correctly provide a high definition measurement of motion and vibrations in the ground. | NA | N/A | N/A | Errors can be found in the model output resulting in a reprocessing. Data quality dependent on input observations (which can improve with new satellite instrumentation) and climate model. | The (resultant) data quality depends not only on the model (satellite data retrieval algorithms and flux model) quality but also on the (input) data quality. | This "knowelege-based" (can also be called as "value-added") service has ensured users to acquire the "right" (adequate and needed) data products. | |||||||||||||||||||||
14 | Visualization | Outputs are displayed in map form on large (approximately 20 foot by 8 foot) screens. | Visualization is a key feature that is desired but not mature in MUSTANG. This is partly because the key goals of MUSTANG is to produce the numbers and make them easily accessible. Visualization is a key feature of seismic data quality analysis. As a result, many traditional approaches are currently in use. Gathering metrics into excel for making plots is commonly used. Visualization of waveforms and noise images is very common. Measuring diurnal patterns to noise is a common visualization. None of these are seamlessly integrated to MUSTANG yet, but are easily within reach. | Visualization used to identify phenomena (e.g. the global monthly mean might not reflect two equally different and extreme phenomena) | Visualization heavily used to spot measurement differences and signatures | Visualizations were used to verify seasonal pattern of the aggregated meteorological parameters, for data exploration (i.e. scatter plot between disease outbreak and meteorological parameters to assess initial or existence of relationship), to illustrate disease prediction and visually compare with observed data. | CREATE-IP will be providing multiple visualization services, including a WMS tool, UV-CDAT, and ArcGIS. | Visualization may mainly serve research purpose so far. | The visualization tool, e.g., the powerful Giovanni, is accessible for users to acquire quick and informative views of their interested data variables/products. | ||||||||||||||||||||
15 | Compute(System), storage, networking | Store the MERRA reanalysis data collection in HDFS to enable parallel, high-performance, storage-side data reductions; Manage storage-side <driver, mapper, reducer> code sets and realized objects for users; Deliver end-user and application capabilities through the NASA Climate Data Services API. | 10 VMs (ea. 8 core, 24 GB RAM) locally, (15 VMs, ea. 8 core, 24 GB RAM) remotely. NetApp filer with ~3 TB allocation, projecting need to ~5TB. NetApp filer replicated at remote site. Servers at each site mount common file system for code and configuration deployment, each has separate home FS. 10 Gbit load balancers and outbound network pipe between local and remote site. VPN connectivity (at much reduced bandwidth) and external connectivity for public HTTP access. | Remote Data Servers: Algorithms such as convolution over sensor point spread functions and spectral response functions are run on them | Each reanalysis project has access to their own computer. Each reanalysis project has local access to their own data but would need to download and format the other reanalysis datasets. Issues associated with moving data between the US, Japan, and Europe. | Datasets produced by scientists using local workstations, then transferred to GES DISC and distributed by GES DISC. | The "Bundled" and the original data variables/datasets are distributed at GES DISC. | ||||||||||||||||||||||
16 | Specialized Software - Current technologies required to perform research - Is there a specific tool, implemented or commercial, that is being used - Software needs should be identified to understand compatibility with available Data Analytics Tools and Techniques | Cloudera MapReduce | Job/Resource coordination using a master scheduler with listeners on remote machines. Analytics carried out using a statistical system like R, central data storage using an SQL database, job state persistence centralized on SQL database. All metrics transactions occur through HTTP web services (read/write). Also implementing a persistent queue (RabbitMQ) to control metrics capture flow from internal and external sources. | New research: N/A | In-house database of remote sensing products where user can retrieve a time series data for any administrative region (country, province and district) or any user-specified rectangular area. Options to average data into weekly and monthly are also available. Mathematical and statistical modeling (i.e. regression, neural network) were performed using Matlab and R software. | Based on climate model code, e.g. GEOS-5. Each reanalysis center has their own code but some are based on the same internal physics code. | Algorithms or/and models for retrieving surface air humidity and computing surface turbulent fluxes, respectively, are required. | Currently, this data service framework is developed on top of existing data services at GES DISC, such as Mirador (search engine), Giovanni (visualization), OPeNDAP, and data recipes. It also involves other popular data tools, such as Panoply, GrADS, IDL, etc. A "Vircual Collection" concept for "Data Bundles", which would involve more computer-driven, machine-learning and software-developing, has reccently been brainstormed at GES DISC. | |||||||||||||||||||||
17 | Current Data Analytics tools applied | Cloudera MapReduce | All original statistics in MUSTANG originate from processing in R. There are almost 50 different metrics being generated for each sensor per day. Analytical processes that occur after the fact take place in R, Matlab, Excel, and seismic viz tools like PQLX. Automated characterization of data quality or anomalies is only now being explored. | MIIC II Framework Uses: Extensible Markup Language (XML) for inter-callibration plan (the event specifications are inserted into this) Open-source Project for Network Access Protocol (OPeNDAP) (event inter-calibration algorithms are executed on remote servers using OPeNDAP server-side functions prior to delivery of the data to the instrument teams for further analysis) | Creating common sampling (time series), gridding. Accounting for temporal and instrument effects. Applying MERRA Sampled like AIRS with Quality control | Regression Modeling, Machine Training and Prediction, Neural Network | Comparison of anomalies, i.e. departures from the climatology (one year's average of the model's 30+ years of data). Analysis of "innovations", the correction applied to the model data to bring the model back to the observations after a set number of model timesteps. | Data format convertion -- from binary to HDF-EOS5 | |||||||||||||||||||||
18 | Data Analytics Challenges (Gaps) | Identifying known data analytics challenges, roadblocks, areas needing attention to accomplish goals | Data sets are very large, so some complex data accesses can be slow when conducted via web services. Unsure of the veracity of metrics algorithms, so expert analysis of hits vs. misses must be carried out with each and every metric to check for proper characterization, correct numbers, proper tuning, and appropriate data selection. Visualization and exploration tools are lacking in MUSTANG (at this stage) to be really effective at analytics. Complete acquisition of data metrics can be difficult due to numerous errors and failures that can happen in the distributed compute pipeline. Detecting these gaps and visualizing data coverage are both immediate concerns being worked on. | Meteorological Data and Processing • Changes in or heterogeneity of: location, formats, algorithm, availability (data continuity) • Data products validation Uncovering patterns & modeling • Choice of mathematical and statistical models • Each model has assumptions such that results and prediction may need to be appropriately interpreted • Parameter constraints and prediction validation | Data volumes, particularly for hourly data. Download times. Format differences. Differences in the definition of the variables by each center. | Producing new sets of massive data products containing valuable physical quanties based on/using several sets of exisitng massive data (satellite observations or model reanalysis) | To develop a Virtual Collecion of Data Bundles. | ||||||||||||||||||||||
19 | Type of User | Taken from the Use Analysis Study, types of user performing use case. Earth science data user classes, who might be performing data analytics, include: - Public - interested user of no or limited scientific skill - Graduate student - person of moderate to high skill at a university or college working towards an advanced degree - Production Centers - large organization that handles/processes vast quantities of data - Science Team - group of scientists focused on a specific area of study or on a specific instrument type, can include cal/val scientists - QA/Testing - developers or scientists using data to test software operation or to determine quality of a product, can include cal/val scientists - Data Analyst - person using NASA data to perform a specific analysis. - Domain Scientist - person using data to do research and publish within a discipline, comes in with some expertise in using the data - Interdisciplinary Scientist - person using high-level data products from multiple sources - Operational User - Data analyst or tech using data for operational support (applications) and emergency response - Assimilation Modelers - persons or groups that routinely obtain vast quantities of data for incorporation into models, can have operational needs - Government Agency or Private Think Tank - a group of people or individual researcher using and analyzing data for decision or/and policy making - Decision Support Systems - computer systems that routinely incorporate data into a system that a less knowledgeable user can use for deriving information | - Graduate student - person of moderate to high skill at a university or college working towards an advanced degree - Production Centers - large organization that handles/processes vast quantities of data - Science Team - group of scientists focused on a specific area of study or on a specific instrument type, can include cal/val scientists - QA/Testing - developers or scientists using data to test software operation or to determine quality of a product, can include cal/val scientists - Data Analyst - person using NASA data to perform a specific analysis. - Domain Scientist - person using data to do research and publish within a discipline, comes in with some expertise in using the data - Interdisciplinary Scientist - person using high-level data products from multiple sources - Operational User - Data analyst or tech using data for operational support (applications) and emergency response - Assimilation Modelers - persons or groups that routinely obtain vast quantities of data for incorporation into models, can have operational needs - Government Agency or Private Think Tank - a group of people or individual researcher using and analyzing data for decision or/and policy making - Decision Support Systems - computer systems that routinely incorporate data into a system that a less knowledgeable user can use for deriving information | Primary user is the network operator: the individual or team of individuals responsible for the deployment, maintenance, and data capture of seismic instrumentation. Though highly technical in their field, they are not generally software people, so the tools can address technical topics, but must be easy to use and 'speak their language'. Secondary user is the scientist or graduate student studying seismic data and wishing to get an assessment as to the quality of the data they are looking at. In the future, we will look at supplying tools to help filter data acquisition based on certain criteria of quality. A final example of user is the data center analyst, who is charged with 'remote monitoring' of stations for problems or data dropouts, as well as assessing the state of recorded data in the archive. The analyst can act as a lookout for the network operator to look into a detected problem, but also serves as a curator of archived data to ensure what is stored is of the best quality. | Science Team | Public user with significant scientific skill, Graduate students, Science Teams, Data Analyst, Domain Scientist - person using data to do research and publish within a discipline, comes in with some expertise in using the data, Interdisciplinary Scientist, Operational User, Assimilation Modelers, Decision Support Systems | Science Team | Assimilation Modelers | Domain Scientists, Interdisciplinary Scientists (Note, above description is related to Domain Scientist usage.) | Research Scientists | Research Scientists, Applications Scientists, General Publics, and Students | ||||||||||||||||||
20 | Science Research Areas | NASA Earth science research areas (http://science.nasa.gov/earth-science/focus-areas/) | Air-sea Interactions. Energy and Water Cycle. P-E (Fresh Water). Monsoons. etc. | Hurricanes. Volcanos. etc. | |||||||||||||||||||||||||
21 | Societal Benefit Areas | GEOS (http://www.earthobservations.org/geoss.php) or NASA Applications (http://appliedsciences.nasa.gov) | Climate Changes. | Disasters. | |||||||||||||||||||||||||
22 | Potential for and/or issues for generalizing this use case (e.g. for ref. architecture) | Unclear. This is a massive anlysis that very for other parties are likely to duplicate. However, this use case illustrates ways to keep the data close to the processing location in order to avoid I/O issues. | Generating metrics to ensure data quality and instrumentation health can probably be viewed as applicable to many forms of instrumentation data gathering efforts. | Same/similar issues arise when inter-comparing any heterogeneous datasets | Tools used can be applied elsewhere. Need to determine where | Building an infrastructure to allow scientists to run similar workflows on the multiple variables from multiple reanalyses. Could be generalized to other ESGF projects as the data formatting is the basis of CREATE-IP | May not be practical to build an infrastructure that may not become popular to be followed by other scientists, but some potential issues may deserve our attention and further discussions such as 1) The Rice Cooker Theory, 2) Producing new (more) "big" datasets based on existing "big" datasets, how do we leverage and optimize such kind of "post-processing" productions, especially when they are inevitable for the Earth science research purpose? | This vlaue-added service of "Data Bundles" could serve as a good model for other NASA DAACs or non-NASA data centers to consider if they have not developed a similar one of their own. Or, this propotype service may develope into a cross-DAAC collaboration or implementation. | |||||||||||||||||||||
23 | More Information (URLs) | Project managers: John Schnase: Senior Computer Scientist in NASA Goddard Space Flight Center’s Computational and Information Sciences and Technology Office (CISTO); Dan Duffy: NASA Center for Climate Simulation (NCCS). http://gmao.gsfc.nasa.gov/merra/ | (1) http://ds.iris.edu/ds/nodes/dmc/quality-assurance/ , (2) http://www.adv-geosci.net/40/31/2015/ , (3) http://service.iris.edu | https://earthdata.nasa.gov/our-community/community-data-system-programs/access-projects/multi-instrument-inter-calibration-miic-framework http://clarreo.larc.nasa.gov/2014-01STM/Wednesday/Currey_MIIC_Framework_CLARREO_STM_123013.pdf | Examples: http://nacp.ornl.gov/MsTMIP_products.shtml http://power.larc.nasa.gov/solar/publications/solar2006_A218.pdf | NASA GES DISC, Thomas Hearty (thomas.j.hearty@nasa.gov), et al; | NASA Global Change Data Center (GCDC), Radina P. Soebiyanto (radina.p.soebiyanto@nasa.gov),Richard Kiang (richard.kiang@nasa.gov) | http://esgf.nccs.nasa.gov Precursor to CREATE-IP: https://earthsystemcog.org/projects/ana4MIPs | http://disc.sci.gsfc.nasa.gov/measures/documentation/Science_of_the_data.GSSTF3.pdf | http://disc.sci.gsfc.nasa.gov/recipes/?q=recipes/How-to-Obtain-Data-for-Conducting-Hurricane-Case-Study | |||||||||||||||||||
24 | |||||||||||||||||||||||||||||
25 | |||||||||||||||||||||||||||||
26 | |||||||||||||||||||||||||||||
27 | |||||||||||||||||||||||||||||
28 | |||||||||||||||||||||||||||||
29 | |||||||||||||||||||||||||||||
30 | |||||||||||||||||||||||||||||
31 | |||||||||||||||||||||||||||||
32 | |||||||||||||||||||||||||||||
33 | |||||||||||||||||||||||||||||
34 | |||||||||||||||||||||||||||||
35 | |||||||||||||||||||||||||||||
36 | |||||||||||||||||||||||||||||
37 | |||||||||||||||||||||||||||||
38 | |||||||||||||||||||||||||||||
39 | |||||||||||||||||||||||||||||
40 | |||||||||||||||||||||||||||||
41 | |||||||||||||||||||||||||||||
42 | |||||||||||||||||||||||||||||
43 | |||||||||||||||||||||||||||||
44 | |||||||||||||||||||||||||||||
45 | |||||||||||||||||||||||||||||
46 | |||||||||||||||||||||||||||||
47 | |||||||||||||||||||||||||||||
48 | |||||||||||||||||||||||||||||
49 | |||||||||||||||||||||||||||||
50 | |||||||||||||||||||||||||||||
51 | |||||||||||||||||||||||||||||
52 | |||||||||||||||||||||||||||||
53 | |||||||||||||||||||||||||||||
54 | |||||||||||||||||||||||||||||
55 | |||||||||||||||||||||||||||||
56 | |||||||||||||||||||||||||||||
57 | |||||||||||||||||||||||||||||
58 | |||||||||||||||||||||||||||||
59 | |||||||||||||||||||||||||||||
60 | |||||||||||||||||||||||||||||
61 | |||||||||||||||||||||||||||||
62 | |||||||||||||||||||||||||||||
63 | |||||||||||||||||||||||||||||
64 | |||||||||||||||||||||||||||||
65 | |||||||||||||||||||||||||||||
66 | |||||||||||||||||||||||||||||
67 | |||||||||||||||||||||||||||||
68 |