USS Carl Vinson, in port at Bremerton, WA, July 2000
|
Experiment Overview
|
|
Experiment Overview
|
|
The overall goal of the 10-13 July 2000 observations aboard the USS Carl Vinson (CVN 70) was to develop and test a methodology to observe how shipboard Navy weather forecasters produce the METOC information that supports the air strike mission. During the experiment described here, the Carl Vinson was docked pier side at the Bremerton Naval Shipyard while she was undergoing maintenance repair. Previous Human-Systems Information Interaction studies had occurred in a very controlled environment. The Carl Vinson study was meant to address the issues of the real-world forecaster who works in a constrained space and has limited time and information sources. The intent was to conduct observations during the Carl Vinson's work-up prior to its next deployment, and study the forecaster in near-combat conditions.
|
|
Experiment Narrative
|
|
Six teams of forecasters and technicians were observed building a strike briefing for simulated air strikes on Port Angeles, WA and Whidbey Island Naval Air Station. The teams were formed from the meteorology and oceanography division (i.e., the OA division) aboard the Carl Vinson. There were three different forecasters (AG-1) and eight different technicians (AG-3) employed in different teaming combinations. The observations were taken in their normal operating environment within OA division spaces on the third deck in the island of the Carl Vinson.
During the observation periods the OA division also used the experiment scenario for training so any time-critical or performance results from the study should be considered carefully. On the other hand, the OA division felt that conducting realistic scenarios this early in the forecasters' and technicians' training was highly beneficial because it would help them work up to deployment.
The forecasting teams were handicapped because the new NITES 2000 system (Navy Integrated Tactical Environmental Subsystem) had just been installed with minimal training. In addition, this study was conducted in an unclassified setting and the time required for downloads of the TAMS/RT Web pages from NPMOC (San Diego) impacted the forecasting. However, the ever resourceful Carl Vinson OA division successfully ran JMV, WinEOTDA, AREPS, SLAP, SMQ-11 (satellite), and normal Web browsing to complete air strike briefings at the end of two hours.
The Carl Vinson environment was very different from that in the San Diego study. More technicians were involved, requiring more coordination. It was also clear that forecasting was a continuous process and the specific air strike components were better integrated when the forecasters had this continuous process already in place. It should be noted that an artificial experiment environment will impact the time to produce the strike forecast.
While there were several analyses conducted, we observed that the use of TAMS/RT improved when high-level guidelines were provided that specifically called for the use of a mesoscale model. However, while local observations were used to validate the mesoscale model, forecasters still needed to consider the relationship of the large, synoptic scale initialization to the mesoscale models in order to properly forecast using the mesoscale model.
|
|
Notes and Lessons Learned
|
|
Observation sessions were partially isolated from interruption from normal OA business operations. OA division personnel understood that training and observation were taking place during certain hours, and either did not enter or restricted their activity in the OA space during those times. However, interruptions from organizations outside the OA division proceeded as normal. For example, all ship-wide announcements could be heard in the space.
The number of observation personnel was limited. No more than two observers were feasible. Observers could not always be in position to see the computer screens and hear the participants. Observers sometimes obscured the cameras as they moved around the space.
|
|
Comments on Methodology and Procedures for Future Strike Scenarios
|
|
Carl Vinson personnel should be consulted in the development of future strike scenarios. Although they were generally comfortable with the scenario, there were some scenario elements that were controversial. Video and audio recording using camera and screen converters is technically feasible in the spaces.
Some additional considerations are:
|
|
- Improved clamping solutions to secure the cameras in place will be necessary for at-sea studies.
- It will not be feasible to mount cameras to take screen shots of individual computer screens.
- It will be feasible to use multiple camera/VCR-converter systems to record screen activity. This approach did not significantly degrade the video output of the NITES computers. Vinson personnel accepted this approach.
- An integrated system should be developed that combines the camera/VCR, converter, microphone preamp, power supplies, and cabling.
- Cameras/VCR tape changes can be difficult given some of the locations of the cameras/VCRs. Changing the tapes on a bottom-loading camera was especially time consuming and may not be feasible once the cameras are mounted more securely.
- Screens that can rotate completely are critical for adjusting camera angles and adjusting and monitoring screen capture.
- Dynamic pan and zoom of the screen converter may be an effective method of capturing details of the screen. This was tested in the last session.
- Background noise from the SMUC-11 satellite fans masks low conversations picked up on area microphones, unless the microphone is very close to the speaker.
- User interaction tracking software (e.g., STARR from iopus.com) can only be used if prior approval is obtained from Vinson and NITES ADP.
- Power outlets for the observation equipment are sufficient.
|
|
Checklist for Strike Support
|
|
A checklist did not seem applicable to forecasting for air strike aboard the USS Carl Vinson during our testing. The reasons for this were both theoretical and practical. Theoretically, checklists need to be based on an established routine that is well understood. Part of our goal was to uncover those routines; we believe, however, that forecasting is highly individualized and may not be suitable for a checklist. Practically, there were limitations to the systems and analyses that the OA division of the USS Carl Vinson was ready to take on.
Therefore, we modified the test procedures to address the following goals:
|
|
- Consider use of the mesoscale model, not the global atmospheric model
- Consider local terrain more carefully
- Make the TDA inputs consistent with the weather forecast
|
|
Overall, the goal was to rely exclusively on the smaller scale atmospheric models that can result in improved forecasting. We tested for these goals by providing a series of critical environmental considerations prior to the test scenario. This was similar to the checklist, but was not structured as a checklist.
Based on preliminary observations our recommended strike forecasting guidelines included:
|
|
- Identifing the major weather events and large scale weather conditions
- Validating the synoptic model against observations
- Validating mesoscale model for target area
- Using validated mesoscale model (TAMS/RT) to construct target forecast
- Considering local terrain influences such as convergence zones and mountain rain shadows
- Ensuring consistency between the TDA and the forecasted weather
|
