RAINEX Project Site Survey Trip Report

AOML/HRD, UM/RSMAS – Aug 30 – Sep 3, 2004


Trip participants – Jim Moore (JOSS), Stacy Brodzik (UW)




We met with several members of HRD in preparation for RAINEX project to take place in August/September, 2005.  We were not able to talk with everyone on our list due to the deployment of personnel in support of operations related to Hurricane Frances.  However, we gathered quite a bit of information and were able to see how HRD operates during a hurricane event.  There was an obvious shift of focus from scientific curiosity about the hurricane to one of preparing families and structures for landfall.


We also spent some time at RSMAS discussing requirements and support for RAINEX science and operations.  We met several key players and agreed to house the RAINEX Operations Center in the U. of Miami complex across the street from AOML.




Key players were identified and visited when possible, and an email list of them is being compiled.


P3 Data – raw data archived on computer but not available online; Nick Carasco is currently setting up online data order form


            1. Types

a.      Flight level – 10 second or 1 minute data

b.      Radar -- lower fuselage (LF) and tail (TA)

c.      Cloud physics

d.      Dropsonde

2. Getting NOAA Lower Fuselage (LF) radar images from NOAA P3’s to RSMAS ops center in near-real-time

a. Capability has been demonstrated with 300 baud link

            b. New hardware has been installed on planes but capability is not

    functional at this time

            c. Moore mentioned that a faster link was used in BAMEX.  (See comment in general observations below)

            d. The interpretation of the LF data in real time is also challenging.  Providing some important documentation on each image (e.g. aircraft orientation and heading, antenna tilt angle) in addition to other aircraft flight track information will be key to providing a useful product to the NRL P-3.


3. Getting data post-mission data from planes to RSMAS ops center

a. Normally not done in a timely manner

            b. Products can be derived on board if we define what we want

            c. Radar, cloud phys data (on 4mm tapes) -> lead scientist (HRD)

            d. Flight track data (on CD) -> flight director (AOC)


G-IV Dropsonde Data – tasking done by NHC but HRD personnel perform QC before data gets to NHC for analysis


NOAA-Port Data – Near-real time data; publicly available; need to know headers.  All but the MinOb data are available via the NCAR LDM services (these are available at RSMAS).  They will be seen in the (*.RECON and *.UPP [for dropsonde data] in the LDM data stream.

1.      MinOb – 1 minute P3 flight level data (This likely shows up via the AOC ASDL (aircraft satellite data link at 100 baud or so).  Access to this data stream will be a key component of A/C coordination.  These messages only come through about once every 30 minutes.

2.      Vortex – location of center of eye, diameter of eye, etc (URNT12 and (URNT14)

3.      RECCO – Message contains time, lat/lon, pressure altitude, flight level wind direction/wind speed/temperature/dew point, and either SLP or geopotential height for some level.  This information is typically only provided at turn points and can be 15, 30 or 60 minutes apart. There can be other information such as turbulence, instrumentation, and present weather.  The optional part often includes sea surface winds (URNT10 and URNT11)

4.      Dropsonde – Mandatory and significant levels (UZNT13)

5.      Lower fuselage composite (not currently functional; talk with Joe G)


Surface Wind Analyses using HWIND (Mark Powell)

1.      HWIND is Java-based (applet version is under development)

2.      Synthesizes winds data from multiple sources (e.g. satellite, aircraft, surface buoys)

3.      User decides what sources to include in analysis

4.      http://cat5.nhc.noaa.gov -- Users Manual, Installation Instructions, etc

5.      Need username and password to access necessary data files


Model Data (Sim Abrasam)

1.      Operational Global Data Assimilation System (GDAS) grids are generated by NCEP and are most useful for post-mission analysis (according to Frank Marks). These grids are received at HRD but we need to check for other sources.

2.      GDAS initializes AVN, GFS, and GFDL models

3.      Consensus track integrates output from GFS, GFDL, UKMET, and NOGAPS

4.      There are about a dozen models used by NHC to consider hurricane motion and speed.

            Note: Need to talk with Abrasam; didn’t have a chance during visit


Satellite Data (Jason Dunion)

1.      Need GOES-E (GOES-12) images in real-time – available from NCAR?

2.      Contact Greg Stossmeister (UCAR/JOSS)

3.      If we want McIDAS files, contact Mark DeMaria or Ray Zehr at

NOAA/NESDIS located at CIRA at CSU in Ft. Collins.

4.      http://cimss.ssec.wisc.edu/tropic


Historical Hurricane Data


Note:  fill appropriate info into “StormNameYYYY”

1.      Mission summaries – including ground tracks and radar composites

2.      Satellite images

3.      Surface wind field analyses (HWIND)

4.      Dropsonde data (raw and processed)

There is generally a several month delay from the time data are taken until

data are made available via the archive.




Daily Planning during Operations

1.      Morning conference call

2.      1230 Briefing

3.      Afternoon – Decision Making and mission selection

4.      Morning – Final Go/No-go decision


General Observations

  1. Due to the difficulty of getting data electronically from HRD to RSMAS, we need to try to get as much of it as possible from other sources.  It will be necessary to create a list of data required for RAINEX from both operational and research sources.  In some cases it may take some time to arrange for collection in real time or after the fact.  This effort will be crucial to providing proper data management support to the project.
  2. Security at AOML/HRD makes conducting routine RAINEX operations in periods other than 0800 - 1630 daily most difficult. Any possible space for RAINEX Operations Center in HRD is very limited and not optimal.
  3. Along with NOAA AOC and NHC, these groups show a well orchestrated planning procedure and process for accomplishing research and operational support tasks.  Adding another facility and aircraft (NCAR/RTF and NRL) to the mix is certainly doable but will require a focus on advance planning and constant coordination in the field.
  4. We have determined that only NOAA-42 has the wider bandwidth capabilities (9600 baud PPP connectivity) as implemented and used during the BAMEX Project in 2003.  NOAA-43 has not implemented this capability, to the best of our knowledge, although it may be available.  This may have some bearing on how coordinated missions are flown with the Navy P-3.  The Navy does have both 9600 baud data and digital satellite voice links.  Note: This approach was the only way to get real-time flight track information from the aircraft (NOAA-42, NRL P-3 and dropsonde aircraft) to the operations center and provide aircraft coordination/severe storm avoidance during BAMEX.
  5. Access to data in real-time from HRD will be a challenge.  As noted above, there may be opportunities to grab data and products or have them sent to us outside the firewall.  We need to pursue several routes for acquiring data needed in the archive.


Frank Marks’ Concerns

            Coordination with Paul Chang

            Coordination of NRL P3 with NOAA P3(s)

1.      Suggested having a dedicated person onboard one of the NOAA P3’s

to communicate with NRL P3

2.      Will NRL plane fly where it’s should?

3.      Will we get radar images to them in a timely manner?


Frank’s restaurant recommendation – Pericone’s – S Miami and S 11th


Actions Items from HRD visit

1.      Griffin – provide appropriate NOAA-Port headers to Jim Moore

2.      Brodzik – get write-up of P3 standard tape format from Neal Dorst

3.      Brodzik – follow up with Griffin concerning feasibility of downlinking LF radar images to HRD/operations center

4.      Houze/Smull – decide which model outputs we want

5.      Houze/Smull – decide if we want satellite images and/or gridded/raw data

6.      Brodzik – explore getting satellite data from NCAR




Key players were identified and visited when possible, and an email list of them is being compiled.


Operations Center

Shuyi’s computer room (19 ft X 15 ft) will be available to us for use during RAINEX.  It currently holds 8-10 Suns and linux PC’s.  They are all on UPS systems.  Power is somewhat unstable and UPS’s are strongly recommended for all computers and peripherals. Still to be determined is who will provide computer hardware dedicated to project.  Shuyi offered one Red Hat linux PC, usage of DLT drives (2 super DLT 320’s, 1 DLT 4000) and an 8mm exabyte drive.  During operations, we will need at least 3 linux PC’s total (with UPS’s) as well as disk space for all incoming data and a 4mm DAT tape drive for reading data tapes from the P3’s.  Joe mentioned that we could get IP numbers from Mike Anderson for whatever machines we use during RAINEX and hook into their network.


First guess at Operations Center residence during RAINEX include:

  1. U of Washington (3 machines, printer??) Zebra et. al.
  2. JOSS (2 machines) Field catalog et. al.
  3. NCAR/ATD/RTF (1-2 machines for ELDORA data handling)
  4. U of Miami (1-2 machines and tape/disk drives noted above)
  5. 3-4 other temporary users


General Info

There may be three different hurricane studies in 2005 falling back to back.  First will be the NASA-funded TCSP program and proposed NSF-sponsored TEXMEX-II running from early June until early August.  Immediately after that, RAINEX, an NSF project, will run from mid-August through the end of September.   It seems prudent to consider setting up the RAINEX Operations Center in June/July to test certain data communication links and on-site Zebra processing and data access that will be required in RAINEX. 


Other Related Comments

1.      Getting data from HRD to RSMAS – there are problems with firewalls at HRD; currently, only way to get data is for HRD to put it on anonymous ftp server outside the firewall; requires man-in-the-loop

2.      Getting P3 radar and cloud physics data post-mission from planes to RSMAS ops center – Joe thinks we can do this electronically

3.      Protocol for uplinking data from RSMAS to NRL P3 – Joe mentioned putting data on RSMAS http server and having planes grab it from there

4.      Format of mm5 output – this might not be a problem since Joe said he can produce netcdf files for direct input to zebra

5.      Next planning meeting – Shuyi agreed to host RAINEX meeting November 15-17, 2004 at RSMAS.  She needs to reserve the room.

6.      The matter of evacuation when faced with an approaching hurricane does present a number of challenges.  The fact is that RSMAS and HRD are both in first level evacuation areas in Dade County, Florida. This means that once a hurricane warning is issued, mandatory evacuation is required.  While we can’t do much about the location of RSMAS (RAINEX operations center) and HRD, we should pay attention to getting personal accommodations that might lessen the likelihood of evacuation quite early.

7.      Access to resources for HRD participation was inferred in several conversations as necessary to get RAINEX-related support in place.  This requires attention by the project very soon.


Summary/Action Items from RSMAS Visit

1.      Moore/Brodzik – get NOAA-Port header info for MinOb, vortex, and dropsonde messages to Joe Tenerelli so he can try extracting that data at RSMAS

2.      Brodzik – send header from existing mm5 netcdf file to Joe Tenerelli so he can write appropriate code to output netcdf mm5 files for use in zebra

3.      Chen – arrange for NRL pilot to fly on NOAA P-3 during hurricane mission this season

4.      Brodzik/Moore/Houze – put together as complete a list as possible of all required data and products (including potential sources) in the very near future; this list should include both real-time and archive data.

5.      Marks/Houze/Moore – discuss issues regarding time and resources to participate in RAINEX and/or support for acquiring data and products needed by RAINEX

6.      Houze/Moore – formulate a RAINEX mission planning strategy