At 5:00 in the morning it can be a little difficult to pick out your bike from about 50 of the same style, 20 of which are the same color! Methinks it's time for some customization. It had rained during the night, but the stars and the last sliver of moon were shining brightly by the time I got moving. One of the researchers on Woja, a small "out island" south of Kwaj, has been having a great time with a small telescope that he brought along. The location is particularly exciting for an astronomy buff because one can see much more of the southern sky than is visible from any part of the continental United States.

On the way to the Operations Center I encountered a somewhat unusual road hazard – a coconut. Later in the morning another trip to the Operations Center would yield my first real encounter with golfers. Yes, Kwaj boasts a nine-hole golf course on the south side of the island, and the weather station, weather radar, and KWAJEX Operations Center sit in the middle of the course. Today was Sunday, so just as at any golf course in the States, tee times were at a premium. Vehicles are expected to stay out of the way, and in any case I don't fancy intercepting a golf ball with any part of my anatomy.

The weather briefing at 6:30 continued to call for minimal rain. There was some hope of continued scattered showers due to a passing upper level system, but as the day wore on the showers quickly fell apart and were replaced by brilliant sun in a blue sky. As a result, the initial call for an airplane mission was dropped. The University of Washington Convair took advantage of the lull to fly past the University of Virginia tethersonde (on Meck Island) to intercompare their various instruments. Once we have a handle on how closely the various sets of instruments match each other in these intercomparisons, we assume that similar relationships hold when we only have readings from one set or the other (which is most of the time!). Unfortunately, the ground equipment for the weather balloon site on Meck Island was still broken, so it couldn't be included in the test.

I took advantage of the lull to work on how best to reconstruct some damaged data files. One of the tremendous advantages of building the data archive during the field experiment is that it forces us to look at the data while we're still here. It has turned out that one of the programs that produces data products from some models of weather balloon doesn't always work correctly, but that if one runs it again (or maybe twice more) it will. We need to re-process the damaged files that we already have, and make sure the crews that use that particular model of balloon know to check the new files that they are producing.

In the evening I made yet another trip to the Operations Center to talk to a post-doc who has been working on the data file problem. We've been banging on the same problem separately because we work opposite shifts and tend to be sleeping when the other person is working. The hours might seem long, but everyone is aware that we only have a few chances to get this right before the project is over. The ride back featured more brilliant stars. Despite the hour, I pulled off the road and tried to pick out some constellations.

George Huffman