In every daily status report, I've noted large variations in the wind variances from certain sites. I've just poked into this a bit. It appears that the variances become large for certain wind directions that generally are site dependent. I've concluded that this effect is from local obstacles (though I haven't been able to completely understand S+ wind directions). For example, at wvally4 I'm pretty sure that the highest variances are for winds over the mortuary building to the north-northeast. River7 also has huge wind variances for the same north-northeast directions, that may also be a building, though this is not as obvious to me.