Code

#historic Davis wind speed
#with CIMIS
#Greg Hirson, 2010

library(cimis)
library(lattice)
library(reshape)
library(ggplot2)

#years = 1990:2009
d = mcimisannual("006", 1990:2009, period = "hourly")
closeAllConnections()

#yearly average wind speed by hour and month
d1 = ddply(d, ~ time + months(datetime) + year,
	.fun = function(df)mean(df$wind_speed, na.rm=T))
d1$time = d1$time/100
d1$months.datetime. = factor(d1$months.datetime.,
	levels = unique(months(d$datetime)))

g1 = ggplot(d1, aes(x = time, y = V1, colour = factor(year)))

g1 + geom_line() + facet_wrap(~months.datetime.) +
     scale_x_continuous("Hour") +
     scale_y_continuous("Wind Speed (mph)") + labs(colour = "Year")

Using the cimis package for R, I put together the following graph. In the graph, the response variable is cumulative degree day - a measure of the amount of heat accumulated throughout the year. It starts in May and goes through July. August will be added as soon as August data becomes available from CIMIS. The heavier red line is 2010 and the lighter lines are the years 2000 - 2009 (see legend).

It appears that is has been the coolest year in the last decade so far. 2004 was also on the cool side. So there you go - 2010 has been cooler than normal.

The code to generate the above graph is included below. The explicit loop was used because for some reason con() was throwing an error when including cimismonthly inside of an lapply().

#coolness of 2010

library(lattice)
library(cimis)

prev = mcimisannual("006", 2000:2009)
prev.s = split(prev, prev$year)

cdd = function(df){
	keep = months(df$datetime) %in% c("May", "June", "July")
	dd = df$avg_air[keep]
	dd[is.na(dd)] = 0
	dd = dd - 50
	cdd = cumsum(ifelse(dd < 0, 0, dd))
	}

prev.cdd =  lapply(prev.s, cdd)

prev.df = data.frame(day = rep(1:92, times = 10),
  cdd = unlist(prev.cdd), year = rep(2000:2009, each = 92))

thismonths = c("may", "jun", "jul", "aug")

this = vector("list", 4)

for(i in 1:4){
	this[[i]] = cimismonthly("006", thismonths[i])
	}

#august from last year until done

this = this[-4]

this.cdd = cdd(do.call("rbind", this))

xyplot(cdd ~ day, groups = year,
	data = prev.df, type = "l",
	panel = function(...){
		panel.xyplot(...)
		panel.xyplot(x = seq(this.cdd), y = this.cdd,
		col = "red", lwd = 2, type = "l")
		},
		auto.key=list(columns = 5),
		xlab = c("day starting May 1"),
		ylab = "cumulative degree day"
		)

For the second tasting, we tasted:

NV Nichelini Napa Valley Zinfandel
1968 Sutter Home California Zinfandel, Lot Number 1, Deaver Vineyard
1972 Mayacamas Late Harvest Zinfandel

1978 Grgich Hills Zinfandel, Alexander Valley
1980 Edmeades Vineyards, Mendocino, Zeni Vineyard
1984 Viano Vineyards Private Stock California Zinfandel

1987 Marietta Cellars Zinfandel, Sonoma County
1991 Orleans Hill Organic Zinfandel, Amador County, Clockspring Vineyard
1998 St. Francis Old Vines Zinfandel, Sonoma County
2003 Marietta Cellars Zinfandel, Sonoma County

The first two flights were interesting. The Nichelini was suprisingly still alive. I thought that the Sutter Home held up best of the tasting - 1968 and it still had fruit and tannins. The late harvest Mayacamas was not my favorite wine - thin and alcoholic, sightly raisiny - consistent with it's labeling but it didn't stand up to the dry wines that surrounded it in the lineup. The Grgich Hills was quite good - lots of fruit and structure, but it had some sort of sulfur problem leading it to have a pronounced boiled onion/beef broth aroma. The Edmeades was doing ok, as well.

Then came the last 5 wines. Oh goodness. This turned into a match-the-flaw-to-the-bottle game. The first wine, the Viano, had a noticeable does of out of place Brett. The first Marietta unfortunately had a TCA/cork issue. The next wine, the Orleans Hill smelled overwhelmingly of bread and butter pickles.

Next came what I have to believe is the worst case of diacetyl in a wine that I have ever encountered. The St. Francis wine was a direct replacement for a popcorn flavored Jelly Belly. It was just plain horrible. The last Marietta was probably the most Zin-like wine of the night. It still had all of its fruitiness, jammy ripe flavors. It also had a lot of alcohol, but that is the style à la mode.

Overall the two tastings together were informative, interesting, and enlightening. CA Zinfandel from the 1960s and 1970s can survive. The 80's were doing ok, as were the 90s. How are the Zinfandels of the 2000s and 2010s going to fare? Will they make it 50 years? I guess we will have to reconvene in 2050 to find out.

Here is a photo of the students tasting.

The first tasting happened last Thursday, July, 22.

The wines from the first week:

1967 Souverain Mountain Zinfandel, Napa Valley

1968 Mirrassou Zinfandel "Third Harvest", Monterey County (Tasting note from Fredric Koeppel here, created almost exactly 25 years to the day before we tasted these wines.)

1968 Louis M. Martini Mountain Zinfandel, California

1970 Souverain Mountain Zinfandel, Napa Valley

1979 Fetzer Zinfandel "Scharffenberger Vineyard", Mendocino

1980 Milano Lane Late Harvest Zinfandel "Scharffenberger Vineyard", Mendocino

1986 Kendall-Jackson Zinfandel "Zeni Vineyard", Mendocino

1991 Rosenblum Cellars Zinfandel "George Hendry Vineyard", Napa Valley

1991 Oak Ridge Vineyards Zinfandel, California

1998 Robert Mondavi Old VIne Zinfandel, Napa County

2001 Steele Old Vine Zinfandel "DuPratt Vineyard", Mendocino

Not surprisingly, the progression in age was the most notable aspect of the tasting. The old wines had the dried leaf, tobacco, musty, neutral flavor while the young wines had a fruity, jammy, oaky flavor.

Less pronounced by just as obvious was the shift in oak usage over time. The older wines (before 1990) had little or no apparent oak influence. As the wines got younger, oak began to play a much more pronounced role. My favorite wine, the Rosenblum, has just enough oak to let you know it was there, but not so much as to be considered a major flavor characteristic. The youngest two wines were almost completely oak dominated.

Another interesting note is the change in acidity over time; the old wines were very tart. We speculated that it may be because the grapes were picked less ripe. Other explanations involved the concept of field blending: When the Zinfandel grapes were harvest, other varieties interplanted in the vineyard were harvested and mixed in with the Zin. A high-acid grape like Barbera may be contributing to the acid structure. We also found a little bit of Brettanomyces going back to the vintages of the 1970's. Though none of the wines were medicinal or fecal, there was some barnyard-leather going on in a few of the wines.

The first week was an informative tasting. None of the wines were so dead that they were unenjoyable to taste. This week we will have 10 or 11 more to try. I will update with another post after that tasting.

What is it, you ask?

1. A monterey jack quesadilla with salsa verde on the inside of the quesadilla;

2. Onions and mushrooms, sauteed with paprika and cumin;

3. Two poached eggs;

4. Topped with more salsa verde.

I'm still thinking about this.

I was first introduced to tahdig at a Persian restaurant in Irvine, CA. We ordered tahdig with a few stews to go on top. Actually, I think we ordered stews with tahdig to go underneath, but I soon learned that the true star of the dish was the rice.

So what is tahdig? Tahdig is crunchy rice created cooking rice hard over heat, resulting in a sort of burned, crunchy bottom layer. Easy right? Add the water to the rice and cook on high the whole time instead of simmering. Wrong. That won't work. Even with the simple instructions on the production of tahdig it took me half a dozen attempts to get it right.

A prerequisite for crunchy rice is... rice. To cook the rice properly, I've found that the bulk water method ("pasta method") is best - bring a whole pot of water to a boil and cook the rice, then drain. This is better than the usual rice instruction, which I call the "risotto method" in which a perfect amount of water is combined with the rice and it is cooked until the water is just absorbed. The bulk method is much more predictable.

The rice is cooked using the bulk method for about six minutes until the rice is cooked but not cooked through - one could call it al dente, perhaps. After that, the rice is drained and put aside. Now, the hard part. How to burn rice in a controlled manner.

At my mom's house in in Los Angeles, she has a gas burner. It is much easier to make tahdig on a gas burner as the temperature control is much better. I have a glass cooktop with a halogen lamp at my house in Davis. So after the hard boil on the rice, I need to cool down the burner. I do this by wiping the cooktop carefully with a wet paper towel. Nothing like using the high heat capacity of water to my advantage, right? After the burner has cooled slightly, I wash out the pot that the rice was cooked in and mix some turmeric (poor man's saffron) with vegetable oil and heat that up on the bottom of the pot over medium high heat. Once the turmeric starts to sizzle, I carefully spoon back in the rice. Now the burning begins. As the rice sits in the oil, it basically undergoes a shallow frying. The pot is covered with the lid lined with a dishtowel to absorb the moisture.

The most important thing is to now make sure the rice doesn't burn and carbonize. I use my nose and the rate of steam escape from the lid to judge. I think it is one of those things that takes practice to know what it is supposed to look like. If I had to describe it, I'd say that the steam comes out at a slow leak and smells fragrant. If the steam starts to smell burned, it's too hot. If there is no audible sizzle, it's too cool. I tend to pull the pot off and back on the heat during the 10-15 minutes of medium-high heat. After these first critical minutes of high heat, the pot is turned down to low and cooked for another 30-45 minutes.

After this is done, if you are lucky, you will find this at the bottom of the pot.

Topped with a lentil stew, a perfect dinner.