In spite of, or maybe because of all the Doppler radar, satellites, hurricane hunter flights, computer models, fancy computer graphics, live TV news and weather coverage including from within the eye not five miles from landfall on Atlantic Beach, highly accurate forecasts by the NHC, and in spite of all the experts and all the rest of us looking at the data, few of us noticed a key aspect of this data and its many implications.

Remember the talking points Friday night and Saturday as Irene was coming ashore in NC: Dry air from south and west had reached and collapsed the storm's eye and little remained of Irene W and SW of the eye; the surface winds for Irene were consistently lower than expected given the HH flight-level speeds and the highest reported winds were down to Cat 1; the highest winds of a hurricane are found to the north and east of the eye; Irene is moving NNE or NE and will soon be out to sea; not much damage near landfall (except the pier) and not much wind or rain reported south and west of the eye. Depending on the site, attacks on the NHC and NWS, on state and local officials, Obama, blacks, the poor, RTC vs state of emergency restrictions on gun, global warming, ...

With landfall 1-2 miles from my sister's house, I started looking at NWS public info summaries of max winds, rainfall S and W along the coast near where I grew up. On first glance, the max winds seemed to have remained fairly low -- too low! And still coming from ENE, E, or ESE -- these values were from hours earlier. The weather stations were offline and had been for 2-5 hours before landfall! Now I was alarmed, particularly since I expected the eye wall replacement was nearly done, with convection wrapping completely around the new eye.

I was using wunderground.com at the time to quickly look at all weather stations from NWS offices and airports down to the simplest backyard PWS that had reported since midnight. Not one of the airport or NWS stations near the coast was still working at 5:00a. About half of those functioning at landfall were part of RAWS, Forest Service stations reporting via satellite. Two important stations also used RAWS: the Cedar Island Ferry Terminal and the NC State Ports location in MHC at the causeway to Beaufort. The latter was used to confirm landfall nearby.

While most Personal Weather Stations (PWS)stopped reporting because their location lost electrical power or internet connectivity, a few continued sending reports. A couple of them were using amateur radio links; a couple of others seemed to be connected using cellphone links. Nearly all the stations still active at landfall had stopped reporting by noon.

While loss of electrical power or of internet access was why most stations stopped reporting, but in several locations, the last few observations showed rapid jumps in the sustained wind speeds and in gusts, coupled with rapid drops in pressure, and with very heavy rain. In at least one of these cases, the station continued reporting temp and pressure long after its last wind or rainfall updates. The time of the rapidly increasing winds and of stations going offline corresponds to the arrival of a very strong convection band. That would suggest that a few "rogue" wind gusts had damaged the instruments at the weather stations.

Whatever the reason, these misleading max winds and not-quite-total rainfall amounts made it harder to evaluate Irene wind or rain patterns, or to convince anyone of the power and danger with Irene.

(Will post more on this later.)