Quick and Dirty Meteorology: Relative Humidity vs Temp Change; Consec Days without Frost or Rain

Several things I've been curious about for a while.

First: I've noticed that on days when it's cloudy or rainy, the gap between the high and low temperature is smaller. This is not surprising of course - water has a high specific heat, and the higher the specific heat of the atmosphere, the more energy must be added or removed to change the temperature a given amount. This is also consistent with areas near the coast with weather incoming from the ocean having more moderate weather, ie, the West Coast is much more moderate than the same latitude on the East Coast.

I used weather data for Pennsylvania, because I was curious about specific things in Pennsylvania, not just this abstract question. I used Weather Underground weather data from Lehigh Valley International Airport Weather Station (KABE). I only looked back the last two years because I couldn't find any way to get historical data other than painfully cutting-and-pasting monthly data from a screen. (If you know a better way, please comment!)

Indeed, I did find that the relationship between average daily relative humidity and the absolute difference between high and low temperatures (moisture to temp range) is negative. The more moisture, the less temperature gap - but the linear relationship is not great, with an R^2 of only 0.0853.

The linear regression shows that the temperature gap closes by about a sixth of a degree Fahrenheit for every 1% increase in relative humidity. The intercept is a 29.4 F temp range, i.e. 0% relative humidity would give us a 29.4 F change.

However, unexpectedly, the same comparison using temperature gap in terms of % difference from max (relative to absolute zero) produces a positive relationship with better fit. And looking at the relationship between high temperature and temp gap, it actually predicts temperature better than relative humidity.

While I didn't do a chi-squared, for what it's worth the average daily temperature gap closes form 20.6 to 15.5 degrees F on days with precipitation.


Second: Stretches of consecutive dry days. California has whole summers without a drop of rain, as a matter of course. When it rains during the summer it's a newsworthy event, and people have unwieldy explanations for what must be happening (hearing rain one summer night, I assumed it was just leaves falling. "That's a lot of leaves," I told myself, "The trees will be bare by morning.") But I was curious about the same question in Pennsylvania. Of course, this is not kept as a running tally - the only systematic way of getting this answer is to look at all weather data and then automate counting days with zero precipitation. For 2019-2020, the longest stretch without rain was 15 days in late September of 2020. During the summer, there are runs of up to 7-8 days without rain. How can people plan anything outdoors in that kind of climate?!? What a hell-scape! I did find this record showing zero precipitation in the month of October 1963, beginning a historic drought. (By comparison, California can have 200+ days of no rain.)

I was also curious about frost-free days, as well as stretches of days entirely below freezing. Of course, it does get cold in the winter in PA, but actually it usually goes above freezing during the day. The growing season - the frost-free season, how many consecutive days entirely above freezing, was 180-some days both years. Both years, the last frost was mid- to late April, first one is late October/early November. Given climate change, it's not surprising that Allentown is shown in a shorter growing season zone than this on this map (you can see it better at accuracyproject.org.)



There aren't many stretches of consecutive days in SEPA when the temperature doesn't get ABOVE freezing at all - maximum stretch in those two years was 5 days. To my mind this is the real measure of the "endurability" of a climate. It's 40 F? Fine, put on a fleece to go running, you'll be fine. And indeed, here there were mostly a few 1-3 day stretches. Compare that just to Fairbanks Alaska's last full winter of 2019-2020, when there were 114 days where the temperature did not go above freezing at all - and that's just because of a single balmy day in late November when it got up to 37 F, and before that it hadn't been above freezing since Halloween.

Willow Creek Preserve, Pacific Union College Forest, and Sept 27-69 - 6:30. by knife

First, and let me be very clear about this, Han Bul is a better restaurant than Soban. If you disagree then I guess you don't really understand authentic Korean food. This means you SS. ZING, HEY-O! (rimshot) (In all seriousness everyone, it's not yet safe to sit in a restaurant - but do support them. Consider getting takeout, or getting a gift card that you redeem later.)

First I stumbled on Willow Creek Preserve by noticing some trails marked on Google Maps that were outside of the (also awesome) Pomo Canyon. I noticed cars parked in a trailhead-like way along Coleman Valley Road and I KNEW you guys were holding out on me, and sure enough I found it.







I'm always struck by the convergent evolution of sorrel and clover, at least in terms of gross anatomy. The former is in the buckwheat family, the latter in the pea family. Per Timetree, their most recent common ancestor was likely 117 million years ago. (For comparison, humans and whales diverged only 96 million years ago.) Sorrel does not fix nitrogen like clover (I assumed there was a symbiosis) but does have a nifty rapid (fast enough to see it move) closure response to direct sunlight, which it's not used to, growing in redwood forests as it does. It's obvious enough why sharks, whales and mosasaurs would evolve convergently, but redwood sorrel is found in the deep shade of redwood forests and clovers in open fields, so this remains a mystery for some keen-minded botanist to solve.

Then I had a little of the ol' patello-femoral recently from a minor acute injury going up stupid Collier Springs Trail in Marin County Watershed land, but still major enough that I don't want to exacerbate it. Was thinking of going up to Wright Peak at Clear Lake, but considering the trail is described as steep and loose, exactly the kind of terrain that caused the injury in the first place, I decided to check out Las Posadas State Forest. Well that's closed. BUT immediately adjacent, I stumbled across the quite-awesome Pacific Union College Forest, popular with bikers, through which the Bay Ridge Trail runs. Even on a holiday weekend-Friday it was still fairly light. It's amazing that I'd never even heard of this place a 45-minute drive from home and it's pretty sweet.







It's interesting to see how each valley over gets progressively drier (Santa Rosa plain, then Sonoma, then Napa, then Pope/Berryessa.) The redwoods and Douglas first more or less stop on the west side of Napa, the east side of Napa looks drier than the west, and once you've over that (into Pope and Berryessa Valley) it's mostly chaparral and a few ponderosa pines and much of it, especially the Blue Ridge on the east side (i.e., the last ridge before the Central Valley) the often eroded-looking dry sandy scrub could pass for SoCal.



After that I gave into my morbid side and found the island (above) where one of the Zodiac Killer's attacks happened. After the dry Mediterranean summer, peninsulas emerge connecting these islands to the shore, as was the case in September of 1969. To take away the fun of figuring it out on your own - it's on the west side of the lake, at the south end of the Oak Shores Day Use Area, near the Lake Berryessa Field Office; nearest mile marker 6.75 along Berryessa-Knoxville Road. It turns out the victims were PUC students as well (I'd never heard of PUC until I ran in their forest mere hours before.) Helpfully, the killer called the sheriff's office that night though boaters had already heard the victims' cries for help (they did not immediately perish) and to further clarify, the killer wrote in marker on the couple's car door "Vallejo 12-20-68, 7-4-69, Sept 27-69 - 6:30. by knife". I've visited the first two locations as well and barely remember what they look like - this one leaves much more of an impression.

East Coast: Where the Trees Don't Lose Their Leaves

I've always wondered where the leaves stay green throughout the year - on the East Coast. Why does this matter? West Coast natives will never understand this, how the combination of the East Coast mentality and continental climate can make you feel trapped by the climate. During a road trip from Philly to Florida at spring break one year, the sun came up right about as we were crossing from South Carolina into Georgia, and seeing the green forest around us, I felt like we'd escaped - like we had cheated the world. It's not even the temperature that can be grueling about a temperate zone winter - it's the colorlessness, the gray clouds and especially the brown trees. A friend from Georgia referred to the Pennsylvania winter's wan sun, clouds and dead trees as the Dead Land. I had the same experience when I moved to California, and drove across the country in December as I descended the Sierras and came into the foothills and Sacramento's eastern suburbs, again there was green. Again I felt like I had come back to life, and I was home.

So I had always wondered: where do the trees stay green?

I started to think I was going to have to learn to program somewhat, and find satellite images from August and February of the East Coast, and then (if I was being rigorous) look at the color value of each pixel and try to correlate that with vegetation. But these maps are already available. (I couldn't find them the last time I did this search.) The first map is from databasin.org. Although weather moves west to east (hence the East Coast being much less moderate), there is still an under-emphasized maritime moderation effect near the coast - for example, the robins don't migrate for the winter in suburban Philadelphia, but one county away in Berks, they do (they come back early February though.) So this map shows how the coastal deciduous zone ends right near the Virginia-North Carolina line, and then hugs the coast, a bit downhill of the Fall Line, maybe 100 miles from the ocean. The pattern continues as it bends around the Gulf Coast, trends a bit north in Louisiana, and then much to my surprise, bends back toward the coast to include Houston. West of that, the mountains and/or dryness of the climate dictate vegetation more than temperature. I can confirm - having flown into Houston at Christmas, I was quite surprised that the trees were brown. (Corpus Christi was still green.) What surprises me most is I think I expected the 40th parallel to dominate more over coastal effects (as the typical dividing line between temperate and subtropical) but this is not the case.

And speaking of the southeast - while you're social-distancing, don't all hike on the same trails! It happened in both California and now Georgia. Come on outdoorsy types. We're smarter than this. And I want more of us to stay alive!



It also confirms what I suspected about Virginia: it's really just Pennsylvania with nastier summers.

Below is a map of the Fall Line from gallivance.net, where the first hills begin. There's nothing like this on the West Coast, where instead of being a passive margin where the land just kind of "gives up" to the ocean, the mountains come right up to the water, sometimes a mile high within a few miles as in California's Central Coast. The Fall Line corresponds to agricultural development historically, the Black Belt, and where the mills were. Almost all of Pennsylvania is uphill from the Fall Line - it's already at the Falls Bridge/Schuylkill Aqueduct, and the Twin Bridges (ie, Roosevelt Boulevard - before Manayunk even!), which is why large vessels don't go upriver past Philly. (Imagine a parallel universe where they dock in Reading. With cruise ships as carriers of disease - norovirus for many years and now COVID-19 - that's probably a good thing.)




Above: from Wiki, the Falls Bridge, aka the Reading Railroad Bridge, and the Twin Bridges at the Fall Line on the Schuylkill in Philly. Below, a typical cruise ship (from Signal Horizon.)



Of course a lot of the non-deciduous forest in the Carolinas is pine forest. This map shows that the vegetation you'll see along I-95 heading south transitions from Loblolly pine to mangroves along the river and patches of chestnut oak.

Earth's Seasons

Redwood Regional Preserve, Oakland Hills

I started in Sibley Volcanic Preserve, then went through Huckleberry and Redwood (on French Trail of course) to Lake Chabot. By early evening the fog came rolling in with a vengeance and I actually got quite cold waiting for the wife to pick me up. I realized that for all the time I've spent in this park, I have very few pictures of it, although the first one below (from early morning run another day, many years ago) remains one of my favorite that I've ever taken.



I originally had some cool close-up images of fern spires here, but my wife hates spores and sporey-looking things and said she'd never read my blog again if I left those images in the post. So I just have a link (below). It's too bad because ferns are cool. They're one of the few plants that can survive on the forest floor, since redwoods are actually quite hostile to most understory plants (block out the light; constantly bombing them with material falling from the tree.) This is why they are NOT good trees to plant at your house - you're guaranteeing lots of expensive tree maintenance and/or house damage 20 years in the future. I have one at home but it's in a pot and when it's big enough I'll plant it away from buildings where it will be happy. I took this picture of the spores under a fern leaf 100% entirely to upset my wife. As you can see it worked. I am a great husband.













Originally here I had a picture of the Narcisse snake dens in Manitoba, Canada, which can now be seen here. Granted this has nothing to do with Redwood Preserve, but my wife is afraid of snakes too, and I'm an unpleasant fellow. Hence my choices.




Above: it's August, and the poison oak is red. Poison oak is worst when it's red; also, when it's every other color. Below: Poisson oak! ZING, HEY-O! Amirite? That's for you statisticians out there, thanks I'm here all week take care of your waitress. Below that, some dummies running around muttering to themselves.





The next three are from Lake Chabot, a) where I once ran a solid trail marathon that kicked off this blog, b) is the closest I've ever come to being bitten by a rattlesnake and c) where I received my hash name in days of yore my liege! Although the air quality improved dramatically this weekend as the onshore flow returned and pushed the forest fire smoke inland, in the last picture you can make out the distant smoke layer.

The Sorted Colors of the United States

Sorted hues from satellite images, from Reddit user OdBx. Average colors of forests around the world and of California here.

San Diego County More Humid Over Time?

Back around 2009 a friend in San Diego told me that he thought he'd noticed it getting more humid in San Diego County since he moved there in the early 90s, and he attributed this to human activity; specifically, people using water for landscaping. I didn't necessarily buy this, partly because like me, he'd moved from back East, and humans acclimatize very quickly to the Mediterranean climate. You get more sensitive to any mild discomfort, and what previously seemed to the climate-beaten Northeasterner like paradise might soon become noticeable humidity.

Or, was he right?

I looked at the Weather Underground historical data for average humidity on August 1 (single day, only because there are no annual or monthly averages, and August 1 because it's warm and it's after June gloom is over so presumably more sensitive.)  First I looked at KSAN, the San Diego airport. For the period 1950 through 2016, there is no change in humidity. But there might not be - the ocean is going to control the climate. What about an inland suburb which experienced growth (and more lawns) in that period? So I looked at Poway. Humidity data is missing for the period 1969 through 1998 inclusive, which is a bit frustrating because if I want to put together a scatter plot showing the relationship to population, that's exactly the period when Poway was growing fastest (from about 7,000 to about 47,000.)

However, the periods 1950-1968 and 1999-2016 are still interesting to look at.


A quick t-test shows that it's a statistical difference, with averages of 51 and 68% humidity respectively. You might actually be able to tell the difference between those two. So is this because of people watering their lawns? Because of some other broader climate change? Or (my bet actually) because of different instrumentation when they started collecting humidity data again? The thing to do would be to look at another inland city that grew over this period (Vista? Escondido?) with complete data, and see if they show the same effect - and then at a town that started small and stayed small. Assuming that there's a 1:1 correlation between population and human-produced moisture in the air from lawns etc., then if you see a difference in trend between the two, OR if neither of them change, then it's not humans watering their lawns. If you do, it supports the argument.

I won't waste much more of your time because here's the data for nearby (also fast-growing) Santee from Weather Underground:



I mean come on. Lots of missing data, but more importantly, this pattern has "different or incorrectly used instrument" written all over it. And I think the people in Santee would have noticed if their humidity suddenly dropped dramatically, and then increased again. So until I can find a reliable source of historical humidity data, the answer won't be forthcoming. (As always, if you know of such a source, please comment.)

I Am a Published Glaciologist

Well okay maybe not. But kinda. First of all if I were actually a glaciologist, I would be more certain about what people who study glaciers call themselves. The only credit I can claim is that I have a photo in Jorge Daniel Taillant's Glaciers: The Politics of Ice; Dr. Taillant is obviously very discriminating in his choices of photographic sources. It's a shot of a rock glacier in Colorado's San Juans that I took four years ago (below; original blog post here).

Blue Ridge/Fiske Peak and the Ohlone Wilderness Trail

That oak is in Ohlone Wilderness, but I did want to document my little trips out to the Capay Valley this spring. Northern California's spring is underrated, and it's already rated pretty highly. I abandoned my initial attempt to get up on the ridge in mid-March in the first quarter mile because the river was out of control, after 2 solid weeks of rain. A few weeks later I came back and it was easily wade-able. Fiske Peak is the northernmost prominent high point along the Blue Ridge, a steep and interesting formation which is little-known to many Bay Area hikers. In some ways it reminds me of Southern California's preserves, even around Stebbins Cold Canyon (very steep eroded sandy ridges, hot in the summer, and inland=drier vegetation). Looking north along the east edge of the coast range you can see that the mountains are a row of several thousand-foot upturned blocks, and this view is hard to appreciate anywhere but up on the ridge. I'd wondered about the complex geology of northern California before. I wish I'd taken my camera!

Fortunately I did take my camera to Ohlone Wilderness. I wanted to get to Rose Peak and back from Del Valle, despite having come down with a cough and fever two days before. Results were predictable. I gave up and turned around at Schlieper Rock, and still feel fried the next day. I are not smrt. At least I got some nice shots.


Above: the trail along the ridge at Fiske Peak. See what I mean? You could be in Orange County. From yolocounty.org. Below: Ohlone Wilderness in May.




Above: mariposa lily, which I had never seen (or noticed). California Native Plant Society on Facebook ID'd it in minutes of posting the picture. Below: Mt. Diablo to the north.

Colors of Forest; Colors of California

Some years ago I noticed the San Leandro Square on satellite images of the East Bay watershed area. So I went back there to see what it was. It turned out to be eucalyptus trees, planted in a right-angled shape oriented almost exactly with the cardinal directions. I never did solve the mystery of what it is (it's not obviously a wind break), but the line of eucalpytus does jump out as darker than the surrounding oak forest. So I decided to take the average colors of an Australian forest (all eucalyptus) and compare it against some others. Below are the average colors taken from satellite images of forests around the world, below:


Top left, redwood forest (Prairie Creek Redwoods State Park, California, USA)
Top right, summer temperate deciduous forest (Sproul State Forest, Pennsylvania, USA);
Lower left, eucalyptus (Wollemi National Park, New South Wales, Australia)
Lower right, tropical forest (Parque Nacional El Mirador, Guatemala)


And just for the heck of it, here are the average colors of a few spots in California (not all forested) below:


Top left, the ocean at Pfeiffer Big Sur State Park, California Central Coast
Top right, Prairie Creek Redwoods again, coastal far-northern California
Lower left, Algodones Dunes, Imperial County
Lower right, the concrete and pavement of San Francisco's Sunset District

To What Extent Does Minutes of Sunlight Explain Temperature?

The list of stupid things I wonder about grows longer still. As I evolved in my complaining about Sacramento's summer weather, I started thinking about to what degree you can predict the daily temperature from the minutes of sunlight. Obviously the relationship can't be 1:1 because the solstices are not the coldest or hottest days on average. But it turns out, for American cities that don't have marine climates, minutes of sunlight accounts for over 70% of the variation in average daily temperatures. You might say, "Of course! What else did you think explained hot summers and cold winters, seasonal volcanism?" It's that I didn't expect the relationship to be so clear, and to be obscured by other climactic factors.

Below please note sunlight-temperature scatter plots for Phoenix, Sacramento, San Diego, and Philadelphia. Initially I expected Sacramento and Phoenix to be similar, and San Diego and Philadelphia (Philadelphia because of humidity and cloud cover). I also didn't expect too tight of a correlation. Temperature data is from 1980-2012.








This was a surprise. Cloud cover and humidity per se don't make a difference, but being downwind from a massive heat sink does. And when you break the temperature cycle in half, and do a separate linear goodness of fit for both decreasing and increasing temperature, the R^2 goes up past 0.9! (Again, for the non-marine cities.)

I even adjusted Sacramento for the differing angle of the sunlight (i.e. a minute of sunlight in January is different than in May, because the sun is lower, and the same power is therefore spread over a greater area; that is to say, watts per square meter is lower in winter than summer.) After all, every place on Earth gets the same minutes of daylight per year, but obviously the higher the latitude, the greater the departure from directly-overhead sunlight, and therefore the less power being delivered, and obviously for this reason higher latitudes are colder. But it turns out that adjusting for incident intensity comes out in the wash, within the same dataset - adjusting for different sun intensity only increased the goodness of fit by about 0.02. There is an obvious difference in the pointiness of winter (Sac and PHX are pointier than Philly, i.e. they start warming after the solstice more rapidly than Philly); this might be explained by difference in cloud cover.

One take home: if you want to identify marine-influenced climates, you could write a program that calculates the R^2 for any city, given latitude (so you can calculate minutes of sunlight each day) and daily average temperature data. The worse the fit, the more there's a massive heat sink nearby screwing up your prediction, i.e. the more marine-influenced is the climate you're looking at.

ADDED LATER: This very nice related NOAA map I found at the Map Porn subreddit. Purpler = sooner (December), green-yellow = later (February).



Temperature data is from average daily temperature archive, University of Dayton.

Moving to Sunnier Places Won't Make You Happy

I checked sunlight hours, average annual temperature, and seasonal temperature standard deviation (i.e. extremeness of seasons) against Gallup well-being data by metropolitan area, and state-level Gallup Healthways happiness data. The scatterplots were a mess (no point in even showing them here) and there was no relationship.

This struck me as very good news actually. Yes, maybe a change might have an effect (a particularly severe winter may depress you if you're not used to it; a week or two in the sunny Southwest might lift your spirits if you're stuck in a dim Pennsylvania winter). But of course there's the infamous hedonic treadmill effect, which is actually good news. Wherever you're living, it's likely the people and the work you're doing there that's determining your happiness, rather than something superfluous like how warm and sunny it is.

Models Show 50% Drop in Sierra Snowpack from Amazon Deforestation

Paper here.

Below is a zoomable map of world forest extent, as well as losses over time. here's a link to a high-res zoomable map showing global forest extent and loss. On the full map the Amazon doesn't look great either, but here I have it zoomed on Southern California. Keep in mind this is the current map, but all of SoCal's forests will have burned by 2040, at current rates.

Where the Trees Are

More here. Man, those people in Nebraska could base a currency on wood instead of gold.