2016 Site Updates

Some links and images are no longer available in our older site updates. Additionally, some information is no longer applicable.

For the latest site updates, click here.

Sonde data from NSF/NCAR's Gulfstream V HIAPER (High-performance Instrumented Airborne Platform for Environmental Research) will begin appearing on our site in real time, when available, starting with tonight's mission sampling around Lester in the Central Pacific. This morning our system was configured to begin allowing the data and we added data from yesterday's first mission around Lester. We will test the system tonight to make sure it works in real time for the next mission into Lester. The National Center for Atmospheric Research, funded by the National Science Foundation, is partnering with NOAA through October 12th while NOAA's G-IV undergoes maintenance.

An August 2nd NOAA mission containing radar data did not get processed correctly due to it failing one of the checks our site does to attempt to verify that the location of a radar scan appears correct. Due to a gap in radar data, our site failed to verify the coordinates and did not display any radar data after that as it would have had the check been passed. We have increased the limits of that check until we rewrite the NOAA radar system, when a similar check will not rely on a gap of data. Additionally, when a non-tasked mission runs past midnight Zulu the data from the prior day is supposed to be moved to the next day. The current radar system does not appear to be working correctly in this regard. The current recon system has both the old radar system and parts of the new radar system. Only the old version of the system is operational, as the newer version is very incomplete. With so much code from both it makes fixing issues very difficult. The newer system has code that is very different, even though they do about the same thing. We are focusing on aspects of the new Cesium display right now but will soon return to updating the radar system.

Cesium has now replaced the Google Earth plugin in our recon system. Additional satellite layers and the ability to map any available daily AXBT data from NOAA aircraft will be added in the coming weeks. Additional testing will continue with the new Cesium display. The Google Earth plugin has long been deprecated, no longer works in many browsers and was to be turned off by the end of the year by Google.

Through the rest of the year we will also work on other aspects of the recon system that have been previously mentioned, as well as improving the display of ECMWF data in the model system.

Various small updates will continue to be made to our recon and model systems. If you have any feature suggestions, please let us know.

ECMWF model data was missed on our site for a few days after Darby crossed into the Central Pacific basin. The bug has been fixed and data should now appear properly. (Due to how the live system works, the missed data will not be available until the storm data is recreated in historical mode at some later date when storm data is recreated)

Live recon mapping using Cesium will be available later this week.

On May 28th we integrated a new feature into the NHC best track and model system. When available, global model data is downloaded directly from the European Centre for Medium-Range Weather Forecasts. (ECMWF) The publicly accessible (WMO Essential) tropical cyclone trajectory data is processed by our site and currently integrated into the data on our site using the model identifiers ECMO, ECME and EEXX, where XX represents the ensemble member number. The ECMWF data available through the Global Telecommunications System (GTS) is not included for a particular run on our site if similar ECMWF data for that run exists in the main model file that our site processes from NOAA. (Andd it has been several years since NOAA has released ECMWF data, which had been added after storms end.) This data is not available for invests. Additionally, due to how our NHC model system works, ECMWF data will not be included as a late cycle model if there is no early cycle model data at the time it is released.

The ECMWF data also contains a little bit of additional data that has not yet been added into the system. The location (coordinates) of the highest wind will eventually be integrated. In time, we plan to make various improvements to some of the model displays, such as being able to toggle the ECMWF ensemble members on and off on the Google Maps display and a folder specifically for the ensemble members in the Google Earth file containing model data. That would make it easier to add and remove all the members at once.

ECMWF data will eventually be made available worldwide in its own system. For now, it only appears for the Atlantic, East Pacific and Central Pacific within our best track and model system.

After some continued updates to improve the code of the model system, we will work again on the recon system. We first will be making updates to the NOAA P-3 radar feature. There will be very little visible change. We need to make some updates to how the data is stored and managed. Next, we plan on updating the recon system to use CesiumJS rather than the Google Earth plugin which is no longer supported and will be turned off later this year. This only applies to the Google Earth plugin, not files available for the Google Earth program. The same KMZ files would be created, it's only the display that will change, from the Google Earth plugin to the pure CesiumJS JavaScript display. We will need to make a slight adjustment to KMZ files since icons must be square in CesiumJS to display properly.

Older model data in our model system has been removed and should be recreated by the end of January. This process takes time. Older invests, since 2003, may not be recreated until later than that. Our site will be installing a newer version of our model system which requires all data be recreated. This will be an intermediate step in our model system update. We will eventually make further improvements to the code, likely in 2017, which will likely require recreating the data again. Other new features may also be added in 2017 to the model system. This intermediate step will allow for a new feature that will then be added in the coming months. The new feature is currently being finished offline and after testing will be implemented on what will then be the newer model system. The new feature also has a broader component that will be added, possibly after the integration with the NHC model system. We will discuss the new feature once it is ready.

The Flash diagrams in the model system have now been updated. If you have any suggestions or find any issues, please contact us.

Next up is updating some of the model system's code which will make way for a new feature in the next few months that will be finished and then tested.

Our system had not yet experienced a tropical cyclone forming in one year and lasting into the next, as was the case for Tropical Depression Nine-C in the Central Pacific. We have now made corrections that should handle the storm properly.

The Google Maps mapping for the recon system will likely not be available until at least the middle part of 2016. Currently I am updating the Flash diagrams for the model system. After that I am working on a small addon to the model system that I will discuss within the next few months if testing goes well. Then I will get back to the recon system. I will try to replace the Google Earth browser plugin mapping in the recon system with an alternative since it is being discontinued. I also need to update the NOAA P-3 radar feature in the recon system to work more efficiently. I hope to have that all done before the 2016 Atlantic season starts. Then I will be working on the Google Maps mapping for the recon system. It might be late 2016 before it is done. I plan on updating the model system in 2017 to utilize better coding and possibly have some new features which I will discuss later in the year as I get more done.