There are several ways of handling data loading in the application, each with pros and cons. We anticipate all of the data loading to fall into one of these categories:
Ceiling and Visibility are available as an observed point variable in METARs, as a forecasted point variable in TAFs, and as a gridded variable from the Ceiling Visibility and Analysis (CVA) product, produced at NOAA's Aviation Weather Center, visible on ADDS, and distributed by the National Weather Service's (NWS) Satellite Broadcast Network (NOAA-Port). The point observations and forecasts are generally available every hour, though they may be upgraded more frequently. The gridded product is produced every 5 minutes and available in GRIB format. Gridded values are continuous across the conus land mass.
Ceiling and Visibility are important to pilots because they determine the airports which are available for takeoff and landing (depending on pilot and aircraft certification) and impact the airspace into which they can safely fly. Pilots need this information in preflight as well as updates enroute.
METAR data is available in 3 forms: as raw text from NOAA, as WXXM from CSS-Wx WFS servers, or as XML, CSV, or binary from the ADDS DataServer.
Airport locations are obtained from the FAA 56-day update FADDS data. They are important to pilots as situational awareness variables, allowing them to determine the proximity of plotted weather phenomena to their departure/destination/alternate or route.
Airport locations could be provided to an app as a static dataset, updated regularly with the application, or as a live service from an operational AIXM service or the ADDS DataServer. Due to the mixed nature of this data, XML and custom formats may be the best exchange model.
Winds are available at various altitudes from the ground (as METARs and TAFs) up through the atmosphere (as modeled winds and FA winds aloft), both as an observed/current speed and direction and forecasted speed and direction. The FA winds are only provided at specific locations, but they are derived from virtual soundings through the model data. Winds determine takeoff and landing direction and safety, fuel consumption en route, and can be used to infer meso-scale atmospheric disturbances. The most important use of winds for pilots seems to be at takeoff and landing, especially for rotorcraft.
Wind data can be obtained as a gridded dataset in MDV from ADDS, in NetCDF from CSS-Wx WCS-RI at NCAR, or as GRIB from the NWS. Winds aloft could be derived from these datasets or obtained in text form from the AWC.
Storms are a supercategory for several aviation hazards, including icing, turbulence, gust fronts, and wind shear. These are usually expressed as an area of varying levels of hazard, which must be interpreted for its impact based on the aircraft type, pilot experience, and passenger comfort.
Storm information typically comes from weather models (RAP, CIP/FIP, GTG, and CoSPA), but may also be human-generated vector products, in the case of fronts. The fronts will be more difficult to process/display. Gridded storm data could be thresholded and/or reduced to low/med/high hazard areas.
NOTAMs are used in flight planning and pre-flight to determine areas where the pilot can't fly or land. These are informational products that should not change often, so, with some exception, they are not needing to be actively presented to the user unless the user seeks them out. No alerting is anticipated for these.
NOTAMs are difficult to obtain. Until recently, they were only on the FAA PilotWeb and TFR sites. ADDS now ingests NOTAMs and makes them available via the Text DataServer. This could be leveraged for this product.
Icing is arguably similar in nature to the "Storms" content above. It is an indication of an aviation hazard. Data are available as grids from an hourly model run including an analysis (CIP) and a forecast (FIP).
Precip Type is of concern primarily near the ground (for takeoff and landing) and for low-altitude flights like HEMS. It affects visibility, icing, and runway adhesion. There are three sources of this information: METARs/TAFs, radar observations, and model analysis/forecast. The METARs/TAFs are more relied upon because they are observation-based or human-generated, but radar and model-based precip type information has the potential to fill in gaps between stations and provide a more temporally and spatially-consistent data.
Radar data can provide insight into summer convection, storm strength, and storm fronts. The most beneficial radar products are the lowest level reflectivity (using either the Base Reflectivity or the NSSL Hybrid Scan Reflectivity) and the total reflectivity through the vertical column (using the Composite Reflectivity). Update rate is the of primary importance and users seemed to like these products as much for their timeliness as for what information they impart about the weather.
The precipitation rate can be an indication of storm strength and is often used in IFR conditions to determine the severity of obscuration. It is a model-derived field present in the WRF-RR.