What is motivation for this new framework in the context of Spenvis as a community resource?
Cleaned data > back to public?
Cross calibration and analyses > requires agreements. Methodology reported in Final Reports
1.1 Stanislav: Proba-V EPT
reasons for PAD exponent differences (Fischer 12; EPT&AP9 20-24)?
(Added by Eamonn: availability of data? complexity of data processing?)
1.2 Alessandro: Pamela
Particle tracing allows separation of trapped particles vs. untrapped/quasi-trapped
K, Phi binning plots: why don't they separate the categories more cleanly?; A: ranges of altitudes is analysed blurring the separation
1.3 Sigiava: SREM data analysis
Construct virtual dataset from Integral SREM
SREM generally gives lower e fluxes at conjunctions with VAP
What parameters is genetic algorithm tuning?: A: Flux bins
Questions on quality of response functions and how IREM comparisons can be used; conjunctions allow rescaling; but it could be energy shift
1.4 Jean-Andre: DEMETER IDP
Wisps observed due to ULF transmitter induced e precipitation; Lighting induced precipitation; drift resonances with E field created by ionospheric winds
Data available; free to use for IRENE if desired; Is ESA going to do it?
Corrections and dead-time treatment not included in dataset on line
1.5 Angelica: IPODE
Very large collection of datasets with associated utilities
How many people use the system? Not possible to say
Is it a formal database? A: It's an archive of data files (CDF)
Are there plans to develop it into a formal database? There are development plans in that direction
How many datasets are public and how many not? Not possible to say
Web interface poss.? Not for data download. Only plot the data
Some datasets provided under CNES agreements (e.g. JAXA) and so no right to make available
1.6 Ingmar: Internal Charging
Galileo EMU but also SEDA, Giove-A, Strv-1d, VAP/ERM
IRENE includes current transmission kernels
Do you have geometry parameters? Yes but they're not public; but design is so simple you can do a simple simulation and its good enough; even 1-D
Have you experimentally tested in an electron beam; differences can be large? A: No but VAP agreement is good so should be OK with the simulation based approach.
1.7 Daniel: ODI Data Sharing
Proprietary Data? A: Data can be flagged as private or put in a separate "private" ODI
Moves to use ODI within IRENE project. Need to capture IRENE needs.
Versioning of data processing? A: Not explicit but can be done by naming
Need a system with easy learning curve; A: it is - has the APIs; can use ASCII
Any plan to harmonise like the Planetary Data System? It was the goal of VIRBO and is effectively being done by CDAWeb.
Anyone can use the ODI to make their own proprietary database system
1.8 DISCUSSION
Data sharing can be risky in that caveats associated with the data must be understood. In some cases its better to keep the data with the originator who will iterate the data with cleaning and reevaluations. Some data processing can be very complex, e.g. the Pamela data and the associated particle tracing to correctly attribute the signal to trapped or untrapped particles. Also Proba-V EPT, CNES SAC you need to understand the history of operation, detector problems, data download issues, etc.
When making data available it is important to provide the associated metadata and reports that describe in sufficient detail the design and calibration. The user must not interpret the data simplistically but consider carefully (e.g.) field of view, energy bins, contamination, etc.
IPR has to be respected. Some data have restriction due to political or commercial reasons and then correct licensing strategies have to be implemented. This is slow, complex and so costly.
When data are cleaned and de-glitched, they should as a minimum be copied to the originator and if the original data were public, ideally the cleaned version should also be made available if only to avoid multiple parallel efforts on cleaning the same dataset.
ODI allows easy construction of private and public databases composed of data in almost any format and with API for many languages. It is freely available...
The low altitude region is important but problematic. The extrapolation of high altitude data to lower altitudes is difficult so in situ data is preferred.
Here is an opportunity to share LEO data usefully with the IRENE project. But need more LEO missions. Although science focus is on higher altitudes, the PEO orbit has advantages of rapidly transiting L space and capturing dynamics.
What is the motivation for groups to share data? Keeping data allows exploitation for publication and maybe also commercial reasons but sharing allows greater scope for joint publication and allows the value of a group's work to be highlighted to funding agencies and so aid sustained or augmented funding of a group.
Can outputs of physics based models be considered as data and treated in the same way? In principle yes, but here the caveats are probably even more important.
Related to the metadata sharing, there are standards in preparation in the scope of the COSPAR PRBEM to standardize the description of response functions along with other data and metadata formatting standards. https://sourceforge.net/p/irbem/code/HEAD/tree/docs/PRBEM_Response_Format.doc