At this evening's meeting we learned that Hurricane Dora is passing south of Hawaii and poses a threat to Johnston Island, about halfway between Kwaj and Hawaii. An evacuation has been ordered, meaning that shipping and passenger service out of Kwaj might get slow for the next few days. Although Dora will pass well north of Kwaj, the remanents of Hurricane Eugene might come close enough to kick off some good rain on Friday.

We are also contending with an outage of the satellite telephones that we use for communicating with the outer islands Lae and Woja. In such a case the backup form of communication is HF radio. The KWAJEX Operations Center has an HF system for communication with the NOAA oceanographic ship Ronald H. Brown, but it's not powerful enough to pull in Lae and Woja. So, about 10 a.m. we gave up on the phones and I went to the Harbor Control center, which does have enough power to reach them. It's not as convenient or as clear as a phone, but we got our business done.

When I say that these and other sites deal with "weather balloons" I am really talking about radiosondes, or "sondes" (the "e" is silent and the "o" is like "aw"). Sondes are small instrument packages about 4 by 6 by 12 inches that contain sensors to measure pressure, temperature, and humidity; a radio; and a battery. The sonde's case is cardboard and styrofoam, intended to be inexpensive and fairly crushable if it happens to hit anything. After all, what goes up must come down! The sonde is checked for calibration and correct behavior, then tied to a large latex balloon that has been inflated with helium until it will lift a prescribed weight. At night a light is tied to the instrument to aid in visual tracking and provide some warning for aircraft. Here at Kwaj they use heavy-duty glow sticks. The whole assembly is released in an open field large enough that the sonde will rise above any surrounding obstacles. Thus the choice to locate the Kwaj weather station on the golf course. After launch the sonde radios back its measurements. The balloon rises until the surrounding atmospheric pressure is low enough that the balloon can't hold all the gas inside it, and it bursts. On a good run this happens at about 20 miles above the earth's surface, although we are the most interested in the first 10 miles. The sonde also provides winds. Older systems, like those used at Kwaj, Meck, and Roi, achieve this by tracking the sonde with a small radar unit. The newer systems at Lae and Woja and on the Ronald H. Brown use small GPS (global positioning system) receivers to provide wind information.

It takes about three days for a technically trained person to learn the procedures and foibles of the sondes and the associated computer systems that record the data. Right now we are training one new crew of two at Roi and two new crews at Kwaj, with another to follow. The crew that went to Lae and the one about to go to Woja were trained last spring. Even with all this equipment and training it is critical for researchers to be looking over the data to detect bad values so that we are not led astray in the research that will be done later.