PROJECT BACKGROUND
One of the first projects I worked on at Creighton (in the summer of 2011) was the creation of an electronic weather mapwall. The scope of this project was simple, we wanted our mapwall to look something like the one in the Weather Center of the Department of Atmospheric Science at Texas A&M. The idea behind this was to create both formal and informal learning opportunities by using past, current, and future meteorological data. This data included recent satellite and radar data, recent METAR data depicting surface weather conditions around the country, and computer modeled representations of future (forecasted) weather. Armed with only the above picture of the Weather Center at Texas A&M, and the National Oceanic and Atmospheric Admisitration's (NOAA) Local Data Manager (LDM), I set out to create a newer and better version of the mapwall, for under $5,000.
PROJECT HARDWARE
For this project I needed to determine what hardware was necessary. The obvious major expense was for the screens. I settled on ten full HD 24" LED screens to display the data (~$200 each), one 48" television as the centerpiece (~$900), and a 15" touchscreen panel to drive the entire system (~$500). This was how the screens were configured:
 |
| Photo Credit: Hallie Dusselier |
After determining how the displays were to be configured, I needed to design the computers to run the screens. I had never built a computer from scratch before, so this took a bit of research into how a computer system works. I decided I wanted as few computers as possible to run the mapwall, so I needed to optimize a system that allowed for multiple independent screens to be run from the same machine. I found a motherboard with an onboard video card supporting two monitors, which contained expansion slots for two more video cards supporting two monitors each. I ended up needing only two computers to drive 12 independently animated displays, which were to be powered by Ubuntu Linux installed via thumb drive, since I had no optical drives on the machines to save money. In the end, each machine (processor, motherboard, video cards, HDD, power supply, and case) cost about $600. With necessary cables and power strips (about $200), I had a total project cost of approximately $4,800.
PROJECT SOFTWARE
Once I built the computing systems, I needed to determine how to display all of the various information on the screens. My only programming experience was with FORTRAN and C/C++. I started by thinking of the logical constructs needed to run some FORTRAN code, and came to trouble in controlling the monitors. Eventually I realized I needed a different language for the program. After doing some research and learning about object oriented programming, I decided I was going to use java to program the mapwall. I had never coded in java, so I learned the language using a few online tutorials and code examples! This was exactly the solution I needed; java allowed me to control each of the six screens on each computer independently. I also learned about controlling a touch screen where events on the screen needed to be recorded as if a mouse was the controlling peripheral device.
Based on the inputs from the touch panel, the code pointed the container for each screen's output to a specific directory where our departmental Local Data Manager (LDM) machine was creating graphics files and automatically copying them over the mapwall machines via passwordless SSH. The files (gifs) in this directory were then animated and these images were shown on the 10 screens. At this point, the user could choose any of the ten screens to be isolated on the large television screen in the center of the mapwall. Using the touchscreen, the user could then control the animation speed and direction of the gifs, using various different buttons and sliders. The user could stop on an individual image (radar, satellite, forecast model output) and discuss what was seen, or slow down the animation to highlight a specific weather phenomenon seen in the data. Once done, the user could select another of the 10 screens, or return to the main menu to display 10 different animated datasets. There were a number of other options for the main display--it could be used as a television to follow local/national severe weather events, external computing devices (iPads, laptops, etc.) could be used to show data that was not programmed into the mapwall.
Based on the inputs from the touch panel, the code pointed the container for each screen's output to a specific directory where our departmental Local Data Manager (LDM) machine was creating graphics files and automatically copying them over the mapwall machines via passwordless SSH. The files (gifs) in this directory were then animated and these images were shown on the 10 screens. At this point, the user could choose any of the ten screens to be isolated on the large television screen in the center of the mapwall. Using the touchscreen, the user could then control the animation speed and direction of the gifs, using various different buttons and sliders. The user could stop on an individual image (radar, satellite, forecast model output) and discuss what was seen, or slow down the animation to highlight a specific weather phenomenon seen in the data. Once done, the user could select another of the 10 screens, or return to the main menu to display 10 different animated datasets. There were a number of other options for the main display--it could be used as a television to follow local/national severe weather events, external computing devices (iPads, laptops, etc.) could be used to show data that was not programmed into the mapwall.
MAPWALL IN ACTION
Once the mapwall was installed, it became a fixture in our the department at Creighton. This room was designed to be open to the students during departmental hours. Students could use the mapwall room to study, or they could use the mapwall to personally investigate phenomena of interest. The personal investigations led to informal teaching opportunities--graduate students would explain phenomena they were studying to undergraduate students.
The mapwall was also used as a formal part of the Atmospheric Science curriculum. One of our faculty held weather briefing for his forecasting and analysis class. These briefings were required of his students, but were opened to the public. Below is an image of a student giving a weather briefing during a formal class meeting, where his personal laptop was plugged into the wall and displayed on the large central television. The other screens were animating data outputs from the North American Mesoscale (NAM) model:
This project was a fantastic experience. I was able to create a deliverable in under four months, with less than $5,000 of capital outlay! Unfortunately...this mapwall no longer exists. It was dismantled as the Department of Atmospheric Science was cut from the University. However, I was granted permission to keep the code I wrote, so it is still alive, if not in physical form.
--David
The mapwall was also used as a formal part of the Atmospheric Science curriculum. One of our faculty held weather briefing for his forecasting and analysis class. These briefings were required of his students, but were opened to the public. Below is an image of a student giving a weather briefing during a formal class meeting, where his personal laptop was plugged into the wall and displayed on the large central television. The other screens were animating data outputs from the North American Mesoscale (NAM) model:
 |
| Photo Credit: Hallie Dusselier |
--David
