CCMC Metadata Registry (CMR)

Go to:

• CMR Home
• View Metadata
• Login

DAFFS (0.37)

Discriminant Analysis Flare Forecasting System

Model Description

DAFFS is a probabilistic near-real-time region- and full-disk solar flare forecasting facility based on magnetic field data, prior flare history, and multi-variable discriminant analysis.  The forecasts arise from probability functions constructed from 2-variable combinations, where different combinations are selected according to performance, separately for each event definition.

Initially developed under a NOAA SBIR, it includes multiple layers of redundancy and potential flexibility, although by default it produces NOAA-like forecasts (see below).   
Redundancy includes providing forecasts based on GONG data if HMI data are not available, then based only on GOES events if GONG data are not available, and finally providing a climatology forecast if no data are available.    Potential flexibility includes numerous options for customization (of threshold, issuing schedule, validity/latency timing, event type).    DAFFS (and separately, DAFFS-G, the “GONG-data-focused” facility) were evaluated in the “Nagoya Workshop” papers (see below).
2023.12: updates coming as DAFFS gets some TLC.

Model Figure(s) :

Model Inputs Description

NOAA near real time (NRT) region identifications and event data, vector magnetic photospheric data (HMI), line-of-sight magnetic photospheric data (GONG).

Model Outputs Description

Probabilistic forecasts for solar flares matching NOAA event definitions: C+, M+, X+ for 24hr validity periods with effectively 0,24, and 48hr latencies.  If HMI data are not available, GONG-based NOAA-region forecasts are issued.  If GONG data are also not available, forecasts based on prior flare history and climatology are issued. 

Model Caveats

Full-disk probabilities (valid for full disk) are calculated from region probabilities.  HMI Active Region Patches are used by default, with relevant NOAA regions associated.  Presently relies on netDRMS/SUMS access. It would not be difficult to forecast for other event definitions but would require re-training.

Change Log

Model Acknowledgement/Publication Policy (if any)

Model Domains:

Space Weather Impacts:

Phenomena :

Simulation Type(s):

Temporal Dependence Possible? (whether the code results depend on physical time?)

Model is available at?

Source code of the model is publicly available?

CCMC Model Status (e.g. onboarding, use in production, retired, only hosting output, only source is available):

Code Language:

Regions (this is automatically mapped based on model domain):

Contacts :

Acknowledgement/Institution :

Relevant Links :

Publications :

Model Access Information :

Linked to Other Spase Resource(s) (example: another SimulationModel) :

Curator: Chiu Wiegand | NASA Official: Dr. Masha Kuznetsova | Privacy and Security Notices | Accessibility | CCMC Data Collection Consent Agreement