Greg

I have some time series of data and I'm doing work in MS Excel. I'm attempting to identify homogeneous periods in time based on several characteristics developed from some metrics I have. The problem I'm having is that the data are pretty noisy, and I'd like to apply some kind of smoothing to the series so I can capture some of the more apparent patterns within the data. They are there, it's clear to me and I can see them, but I'd like to smooth the data to improve my identification of these periods. I know the built-in line fitting and moving average functions in Excel, but they're not really what I'm looking for.

If someone can point me to some technical resources, academic journals outlining the processes in a practical manner, good books, websites, give me a rundown, email me, whatever, it would be much appreciated.

I'm interested in doing restoration work on a mid-80's Buick Skyhawk T-type. Not a lot of these cars were made, and the ones that were made weren't treated very well. There aren't many on the road anymore and a lot of the owners that still have them are hesitant to give them up.

I guess what I'm asking is what I should do to try and find one (and parts.) I've tried the usual, like eBay Motors and other used car search sites but I come up pretty empty-handed. I haven't come across any dealers or used car lots that have them sitting around. I was wondering if anyone knew of a resource I could use to try and find some.

Thanks!

Apologies for the wall-of-text.

Maybe someone can improve my understanding of this process because I don't know if I'm understanding it correctly. By the Clausius-Clapeyron relationship, the atmospheric capacity for water vapor increases by around 7% for every degree Celsius that it warms. I've seen this relationship in atmospheric science textbooks, the IPCC reports, peer-reviewed papers, etc. Now to my understanding this is a cause for what is being referred to as "intensification" in the hydrologic cycle through its main driver, precipitation.

The physical basis for this is that the saturation vapor pressure for the water vapor increases and therefore a larger amount of water can be held in the air. Now when you consider precipitation processes, you need moisture, cloud condensation nuclei and a cooling mechanism. We won't consider the availability of CCN here, and we will assume that at least pseudoadiabatic cooling still occurs when you lift a parcel of air. But in order to reach condensation levels (beyond saturation) you need a relative humidity of >100%. This is where I'm confused. If you define RH = e/es, where e is your vapor pressure and es is your saturation vapor pressure, an increased es will decrease the RH, making it required that the parcel be lifted higher and cooled further in order to reach supersaturation (and for condensational growth to cascade to collision-coalescence processes and precipitation to form.)

So wouldn't that really decrease the amount of precipitation, not increase it?

All I can really guess at is that with the increased actual quantity of water in the air (like your mixing ratio or specific humidity), once the precipitation process begins there is more available moisture to keep the process going with continued lifting, as in orographic effect cases or a very strong updraft. This makes sense when many papers assert that while the frequency of small intensity storm events (say <20-50mm depending on where you define your threshold) won't change, larger storms will occur more often, increasing the overall volume of precipitation (hence "intensification.")

If you need more details, I can provide papers and such that illustrate these assumptions but hopefully this outlines it well enough.

Thanks!