Wednesday, July 29, 2015

Search Challenge (7/29/15): Identify these three insects

Entomologically speaking...

... this Challenge shouldn't be that difficult.  But as we've seen, sometimes these things can spiral out of control and present real challenges.

This summer has been a series of insect encounters. Luckily, few of them have been biting insects (mosquitos, ticks, ants, spiders, etc.), but instead they've been moments of beauty and wonder.

Thing is, these lovely creatures are sometimes difficult to identify.  Even though I like to think I recognize most flora and fauna, I find it difficult to remember which kind of insect is which.

For today, three Challenges:  Can you identify the common and Latin names of each of these?

A few hints to get you started:  These are all local to where I live in California. I list the county name where each was observed.

1.  What's this?  (I spotted this at the Sunol Regional Wilderness, Alameda County, California.)  

2.  What's this?  (Note that it's very similar, but not quite the same as butterfly #1.  This one was found in Santa Clara County.)

3. What insect is making this sound?  Can you figure out not just the kind, but also the genus and species names?   (This was recorded around 6PM on a warm summer's day at the Pearson-Arastradero Space Preserve.) 

Can you figure these out?  

If so, let us know HOW  you found them out.  (How on earth will you figure out what kind of insect is making that sound?)  

Search on!  

Monday, July 27, 2015

Answer: Time zone eccentricities

Once again, the reality is more complex... 

... than I thought.  If you're like me, you thought that there might be as many as 24 time zones.  You probably knew that China was all one giant time zone, and that Newfoundland (Canada) was a bit odd.    

Time zones don't often follow straight lines
But as we know, ofttimes when you start to look into something, things get more complex than you'd imagine.

Let's look at our questions one at a time. 
1.  What time is it in Newfoundland (Canada)?  And why is it different than the time in Prince Edward island (which is due south of Newfoundland)?  

It's pretty straightforward to find out what time it is in Newfoundland--the query: 

     [ time Newfoundland ] 

and learn that the time in Newfoundland is GMT-2:30.  


Yes, it's NOT an integer number of hours different than GMT.  (Unlike, say, the Pacific Daylight Time Zone, where I live, which is GMT - 7 hours; or the Atlantic Standard Time Zone (which PEI follows; defined as GMT - 3).  

Why off by 30 minutes? 

The NTZ entry on Wikipedia tells us that "This time zone exists because of the location of the island and the fact that it was a separate dominion when the time zones were established. The island of Newfoundland lies squarely in the eastern half of the Atlantic Time Zone, exactly three and a half hours from Greenwich. Since it was separate from Canada, it had the right to adopt its own time zone. While the entire province lies west of the standard meridian for a half-hour time zone, 52.5 degrees west longitude, this is also the near exact meridian of St. John's, the province's capital and largest city. In 1963, the Newfoundland government attempted to bring the province into conformity with the other Atlantic provinces, but withdrew in the face of stiff public opposition." 

Ok then.  It fits better with the local time, and the locals like it. (Which also, by the way, gives the Newfoundland TV station the ability to run "Television World Premiere" as a bumper when it shows some new TV show...)  

2. Can you create your own time zone map by finding data that gives the actual boundary lines for each zone?  (If you do make your own map, please share the URL for your creation!  We'd all love to see it.)  
 There are many ways to do this.  I really like the map that CartographerWithoutBorders author Brandon Rishel made. (He has a great blog--highly recommended if you're into maps.)

I ran across this map before I wrote this Challenge and liked it so much that it led me to re-cast his map as the Challenge.  (H/T to Brandon!)  

Time zone map created by Brandon Rishel.
When I asked him, Brandon kindly replied that he got the data from and created it from that.   (In particular: ne_10m_time_zones.shp, from the 1:10m cultural vectors.)  He also said that NaturalEarth is an amazing collection of map data--he's right.  

But that motivated me to look around a bit for other time zone data.  Since I know that Google Earth and Google Maps accept KML (as well as KMZ or SHP files), I did a search for: 

     [ time zones filetype:KML OR filetype:KMZ OR filetype:SHP ] 

Once you do that search, you'll find there area a TON of data sets.  So I downloaded a bunch (trying to look for ones with fairly recent dates), and ended up just importing them into Google Earth Pro to plot them on the globe.  

While there are many, I especially liked the tz_world_mp shapefile  (which was labeled as having ALL of the timezones worldwide, from 2013.  The data source page also says that "The geometries are primarily derived from the fip10s data (itself derived from the VMAP0 data, provided by the US National Imagery and Mapping Agency--aka NIMA), and then augmented with data presented in the pages for the maps of the United States, Canada, Russia and China.

That sound pretty authoritative.  

Once I downloaded the data file and unzipped them, I had a 33.9Mb tz_world_mp.shp file.  (That is, the "Time zones world multipolygon" shape file.) 

All I had to do was to Import into Google Earth Pro and I found I had this... 

What's up with this?  Why isn't it more like Brandon's map, or more like the traditional time zone maps I'd always seen? For instance: 

Time zone map from Wikipedia.

What Brandon did in his map (and what most traditional timezone maps do) is to merge time zones that follow the same time procedures.  When you merge equivalent time zones together, you get something more like the ordinary map.  

But there's still something interesting here when you look more closely.  

Here's a chunk of southern US and northern Mexico.  

Now, here's that same area with the timezone polygons layered on top of it: 

What ARE all of these colors? 

To find out, I zoomed in on Arizona and clicked on the middle of the slate-blue area near Phoenix. 

That popup appeared when I clicked on it.  Okay... NOW what's "TZID America/Phoenix"?  

A quick search for that teaches me that "TZID America/Phoenix" is the "Time Zone Identifier" for that region.  TZIDs are defined as the authority for time definitions by region.  The database, as well as some reference source code, is in the public domain, with new editions of the database and code being published as needed (usually a few times each year).  The files are stored (and available at the IANA FTP server  (Or you can see the complete list of timezone codes at Wikipedia.  Caution: This is a fairly large web page.)  

If you look up these definitions (e.g., "TZID America/Phoenix" you'll find that it's the same as "Mountain Time" (but without following Daylight Savings time).  Here's the entry in the table: 

The first -7:00 is for Standard time, and the second -7:00 is for Daylight Savings time (which means they, Arizona, don't follow Daylight Savings time.  And neither does anything that's colore slate-blue.  

But here's the entry for TZID America/Shiprock (which is the same as America/Denver): 

Shiprock is the central city in the Navajo Nation in northern Arizona and it's colored green above.  Which answers our next question... 

3.  Once you've made that map, what's the strangest time zone anomaly you can find?  

Check out this map (I did a search for [ Navajo Nation ] on Google Maps, which highlighted the boundaries of the Navajo Nation in light pink.)  

But within the Navajo reservation is the Hopi reservation, and within that there's a Navajo enclave... and likewise, there's a second Hopi enclave!  

As you can see from the map above, the Navajo follow TZID America/Phoenix, while the Hopi follow TZID America/Denver. 

Since Denver follows Daylight Saving time (they go back one hour in the fall) and Phoenix does NOT, this means that a straight-line 100 mile walk, it's possible to shift time zones 7 times.  

We won't even start to talk about that complicated area near Gallup, NM.  But you can lookup why TZID America/Ojinaga (that heavy pink line running at the bottom of New Mexico and Texas) has its own Time Zone.  

For a truly brilliant explanation about Daylight Saving Time, TimeZones, and about the Navajo/Hopi time zone eccentricity, I highly recommend CPG Grey's video on the topic.  (H/T to Simon Tong for the link!)   His Arizona eccentricity starts at 5:01 into the video, but the entire video is worth watching.  

Search Lessons

There's always something to learn here... 

1.  The data is out there...  As we found, there are many, many versions of the TimeZone data.  Keep searching if you're not happy with what you've found.  The truth, as they say, is out there.  (And a big thanks to Brandon Rishel for pointing out the Natural Earth data site.) 

2.  For lots of easy geographic applications, Google Earth is your friend.  Note that the Google Earth Pro version is now FREE!  (Hurrah!  It used to be expensive.)  This allows you to import all kinds of files and then slice/dice them for layering and analysis.  

3.  As always, there's usually more to the story than you expect.  I could have written a book about all the things I learned about Time Zones.  What at first seems like a simple concept turns out to be really rich and interesting.  (Which is why it took me so long to write this solution up this week.  Waaay too many interesting things to read.)  

Search on! 

Thursday, July 23, 2015

Search Challenges Revisited

As you've probably noticed, 

once you've spent some time on a topic--like our Search Challenges--you start to notice these things in your daily life.  

Comet tales:  I know a lot of you have been watching the saga of Rosetta / Philae and the story of Comet 67P/Churyumov-Gerasimenko.  As you remember, the lander finally landed on the comet and then fairly quickly went silent.  (Apparently it landed in the shade and couldn't recharge its batteries, so it went into hibernation.  Then, after a number of weeks, the sun came out, it warmed up and got a charge back.  Now it's gone offline again (on July 9, 2015).  With luck it will come back. It's done so before.  Stay tuned.  

But for a while there, it was doing science, sending back data.  Here are a few of the findings about the Comet.  (You can read the blogs about Rosetta, or about Philae and get the latest info.)  

Giant sinkholes seem to have formed by some kind of collapse on the surface.  The pits can be found in just a few regions on the comet. There are small groups of them on both the "head" and "body" of the rubber-duck-shaped comet, but nearly all of them appear in the comet's northern hemisphere.

Cameras on Rosetta's OSIRIS instrument have spotted dust jets shooting out of some of the deeper depressions, but those that are more shallow do not seem to be active.  

Composite slide from ESA.

Mudejar Stars:  I don't know about you, but I keep spotting these everywhere.  Of course, it helps that I live in California, which still has a strong Spanish heritage.  I'm in Pasadena today, and as I was walking to the Cognitive Science conference, I walked past a pretty table... with that star pattern.  I like this not just for the repeated star pattern, but also for the "negative space" between the stars (look at the red plus-sign pattern in between the stars).  

Anode in the street:  Remember our discussion about those things embedded in the street labeled "Anode"?  

I recently found this in my neighborhood: 

And... right next to it... 

And it has THIS on the back of the box, which pretty much confirms what we found in our original research.  

Basically, reading this blog primes you to notice these things.  

And if you notice more about how you search, and learn a few really interesting things about the world, I'm happy--the blog is succeeding!  

Search on! 

Wednesday, July 22, 2015

Search Challenge (7/22/15): Time zone eccentricities?

Time zones... 

... are a 19th century invention, really.  TIME has always been around, of course, and for the largest part of human history, time was defined by whenever noon happened locally.  That's easy to figure out, but as tech advanced, we suddenly needed to define time carefully so (for example) trains could arrive on time.  

Time zones don't often follow straight lines
For quite a while, different train companies had their own "time zone," which was clearly crazy.  (The stationmaster had to keep multiple clocks to know what time each train system thought it was, and when they'd arrive.) 

Eventually we figured out that creating time zones, that is, regions of the planet that all share a single clock time, was a great idea.  

The obvious thing to do would be to create 24 time zones, each defined by a line of longitude.  But then things get wonky:  the longitude lines ignore things like country and city boundaries.  It would be a real hassle to have the west half of your town in one time zone, and the east half in another.  

So.. time zones have interesting boundaries.  I was looking at one such map the other day and noticed that it's extremely strange in places. 

And when things get strange, that looks like a Search Challenge.  

So here are 3 questions for you: 

1.  What time is it in Newfoundland (Canada)?  And why is it different than the time in Prince Edward island (which is due south of Newfoundland)?  
2. Can you create your own time zone map by finding data that gives the actual boundary lines for each zone?  (If you do make your own map, please share the URL for your creation!  We'd all love to see it.)  
3.  Once you've made that map, what's the strangest time zone anomaly you can find?  

I know this sounds kind of difficult, but I suspect it's simpler than you might expect.  

And when you've found that anomaly, be sure to tell us HOW you found it and WHY you think that's the oddest anomaly.  

(While you're at it, can you figure out why it's so anomalous?  Is there a story there?  In my experience, there usually is...) 

Here's one of my favorite pictures from my 2004 trip to Fiji.  It's where the date line crosses through Taveuni.  Yeah, it's Sunday on one side of the sign, and Monday on the other.  

It has to happen somewhere! 

Search on! 

Monday, July 20, 2015

Answer: What kinds of buildings are these?

Three different building shapes... 

How hard can it be to give names to these?  Here's what I did to figure these out... 

1.  What is this kind of building called?  Why does it have this distinctive shape?  

Building Type A: It's located at 47.5529417,-122.3365972   -- What kind of building is it? 

I've seen this peculiar building shape in lots of industrial areas, but never really knew what it was all about.  The Challenge is to figure out what/why/how!   

I searched for quite a while to figure out this one; probably 90 minutes all together.  I tried queries like [industrial building profile] or [shop building elevation] or even [industrial building design] (along with truly hopeless queries like [roof with horns] and [dual ventilation industry]).  

Nothing worked.

Out of desperation I wrote to an architect friend and asked--"What is this?"  He replied, "I have no idea, but I've only ever seen them used for glassworks."  

So I went back to Streetview for that location and checked out the sign on the street.  Notice that building shape in the upper right background, AND notice the graphics on the wall behind the sign... lots of glass bottles. 

Streetview of the corner near the lat/long given.
A quick search for [Ardagh Group Seattle] tells me that they are known for "Excellence in glass and metal packaging."  A ha!  A clue! 

My next search was: 

     [ glass factory roof ] 

which led me to a QA site at the Corning Museum of Glass where someone asked the question: "Old glass factories have a distinct kind of roof design. Is that for ventilation? What is it called?"  

The answer given is (in short) "Robertson Ventilator."  From there it's a short query, [ Robertson ventilator ] to 
 the Robertson company site which explains that these devices are generically called  monitor ventilators.  (Robertson is the company that makes the most popular version.)  That site goes on to explain that "Monitor ventilators provide the necessary capacity to meet the exhaust requirements of extreme heat and fume removal problems such as those associated with the glass and steel manufacturing industries. Internal conditions in industrial buildings producing extreme degrees of heat and fumes often cannot effectively be relieved by installing unit ventilators or by conventional continuous ridge ventilators 

Armed with that information, it's pretty easy to find buildings with these devices. 

Building Type B:  You used to see this distinctive shape everywhere.  There are still a bunch of these buildings out there, often repurposed from their original use.  What are these things called? 

I think most people just recognized this as a Quonset hut.  I'm impressed that so many people just know this.  

But if you didn't know what it was, how would you search for it? 

In my case, the very first query was just description:   [ semi circular building ], which gave me lots of round-buildings in exotic locations, but when I added in the material, like this:

     [ steel semi circular building ] 

you can see I found it pretty accurately. 

I was impressed with all of the web sites that Regular Readers found.  Check out the Quonset Hut: Metal Living for a Modern Age website (and book), or the Seabees web page about Quonset huts.  One of the key advantages of the Quonset hut design was that  it's" simple to erect that anyone who could hammer a nail could set it up. A crew of just 6 experienced men could build a hut in a single day."  

Building Type C:  Here's another common industrial building design pattern.  Why would they make the roof like that?  

This is the original HP building from many years ago.  And when I did my search for this, I got lucky and searched for: 

     [ saw tooth roof ] 

  and found that these roofs are designed with a series of ridges with roofing on one side and windows on the other. The steeper surfaces have windows and face away from the equator to shield workers and machinery from direct sunlight. This roof shape lets natural light into a deep plan building or factory.  (And not incidentally, makes it easy to ventilate a busy working area.  See the monitor ridge ventilators of above.)  

Search Lessons 

In two of the three cases, fairly straightforward descriptions of the buildings just worked out without any problems.  I DID have to modify my query about semi-circular buildings to add "steel" into the query, but that only took a second or two. 

But finding Building Type A was tough!  Why was it so hard? 

Because I had almost nothing to go on.  I didn't know what to call something like that, and without a speciality term, all I had were vague, overly general search terms ("industrial" and "horn-shaped").  

Once I learned that it was a shape associated with glass manufacturing, I had something specific to work with.  So my lesson for this week is really... 

When you can find something specific to your query, use it.  (But be sure it's the right thing, lest you run down a long rabbit hole!)  

Search on! 

Friday, July 17, 2015

A bit of a hint...

I have to admit... 

... that figuring out Building Type A was a little tricky.  It took me about 90 minutes of searching before I found the path that worked.  

But I WILL tell you that between the post and the comments, you have all the information you need.

My hint, though, was the observation I made that the "horns" at the top of the building are an add-on to the building itself.  They're commonly used for these kinds of manufacturing buildings.  

In this sketch, the black horns + the M-shaped piece are all thought of as one unit that sits on top of the building.  

FWIW, I made this sketch in the hopes that doing an Image Search on this would give me the right answer.  But it didn't work out; I got lots of nothing with this.  

On the other hand, now you know: All of the information (key terms and concepts) you need is right there in front of you.  

Search on!  Tell us how it's going for you? 

Thursday, July 16, 2015

Announcement: A new version of will open on Monday, July 20, 2015

As you know, 

my MOOC ("Massive Open Online Course") called has been fairly successful.  We opened it a little over two years ago, and it's had a bit more than 600,000 students go through the materials.  

I'm happy to say that we're launching a NEW and UPDATED version of the MOOC this coming Monday.  

You can register now, and then proceed through the course at your own pace after it opens.  I'll open the first half of the course on Monday, then the second half the following weekend. 

This is what a we call a "soft launch" -- that is, the total number of students in this offering will be around 1000 or less.  We're keeping the number intentionally low so we'll have the capacity to handle issues if / when / should they arise.  

We'll then re-offer it in August and September (and probably forever more after that), but I'll be around to be a TA and answer your questions during the next 3 cycles of the class. 

Now, if you're a regular reader, there's probably not much in the MOOC that will surprise you.  Still, if you're a fan of SRS, you might consider checking it out.  If you score well on the midterm and final, you'll get your own certificate of successful completion of the course (with your very own gold star)!  

I'll be announcing the next several offerings when we get closer in time.  (But think mid-August and mid-September for the next couple of classes.) 

If you want to come along for the ride, register at 

Hope to see you online! 

Wednesday, July 15, 2015

Search Challenge (7/15/15): What kinds of buildings are these?

I have to admit that I'm fascinated... 

... by old industrial buildings.  Maybe it's a side-effect of having grown up in Los Angeles, where there were many older buildings from the industries that abounded in south-central LA.  I saw all of these buildings growing up, and now as they begin to fall out of favor or use, I can't help but see them with a bit of nostalgia.  

But despite having grown up with these buildings, I'm not sure I know all that much about them.  Can you help me identify what these typical industrial building types are, and why they have these peculiar shapes? 

I have three different shapes of buildings, and the Challenge for this week is the same for each type: 

1.  What is this kind of building called?  Why does it have this distinctive shape?  

Building Type A:  Here's the building that started it all for me.  I was driving near the docks in Seattle, saw this building, and shot the picture when I paused in traffic.  (It's at 47.5529417,-122.3365972 if that helps.)  

I'm still looking for a better picture, but you see this kind of building everywhere in industrial zones.  They all have the large double curved shapes growing up from the roof.  What ARE those things?  Why do they have such a distinctive and odd shape?  (If I find a better picture, I'll update the post.)  

(You can click on any picture to see it full-size.)  

Building Type B:  You used to see this distinctive shape everywhere.  There are still a bunch of these buildings out there, often repurposed from their original use.  What are these things called? 

For what it's worth, the second building is about 2 km from my house and was reshaped a bit from its original design to become a garage.  (This photo is from a few years ago--it's still exactly the same, only the cars have been updated.)  

Building Type C:  Here's another common industrial building design pattern.  Why would they make the roof like that?  

The bottom picture is the original HP building from many years ago.  This view can't be seen any more, but for years, the building poured light into the night sky of Palo Alto.  I lived about 1 km from this building for a while, and got to know it well.  As with the others, this is a common pattern you see in many places.  Just curious:  Why such an odd roof? 

As always, please let us know HOW you found the answer.  Your path to success is of real interest to all of the readers.  (So don't just tell us "it's this!" -- tell us how you came up with your successful search queries, and what resources you used along the way.)  

Search on! 

Monday, July 13, 2015

Answer: Data for the Greek referendum?

This kind of question comes up all the time... 

HOW do you find official information sources, especially for events happening outside of your normal information channels?  

The Challenge for this week was pretty easy to state--the Challenge is all about the July 5th referendum in Greece about accepting (or rejecting) the proposed austerity measures for continued financial support to Greece from the EU.  We know all this, but...  

1.  Can you find the OFFICIAL source of the vote data?  (That is, the official government reporting on votes by precincts.)  

2.  Once you've found that data, can you graph or plot it in some way, voting-area by voting-area?  (Ideally, we'd like to see if there were some regions of Greece that were especially Yes or especially No in their voting.)  

 My approach:  

Since I wanted data from the Greece government (and I don't know off-hand what the Greek government top-level domain name is), I tried just limiting my search to Greek sites with this query: 

     [ greek referendum data site:gr ]

And THEN I limited my results to all those within the past month (I used the Search Tools pull-down option to do the date-restriction).  This gave me a few good looking results: 

As you can see, this shows three different sites: 

1. -- you can check out their EN language "About" page and find out they're an opinion polling firm for Greece.  

2. -- this is the Ministry of the Interior for Greece.  (Probably a reputable source.)  

3. -- which advertises itself as a news site about Greece, reporting from a "Left Politics" perspective.  (It's nice to see a newsite that makes its perspective clear!)  

Do these sources agree with each other? 

PublicIssue reports that 61.3% voted no. 

Ekloges (the Ministry of the Interior) reports that it was 61.31% no. 

And AnalyzeGreece reports the same, 61.31%.  In fact, every place I checked reports the same number.  No doubt there.  

Handily, the Ekloges site also includes a tabular breakdown by "Election Districts" (or "Major Regions," if you want less granular data.  

I was going to download the data and do some graphing, but a couple of additional clicks led me to the district-by-district map of voting results.  

That's pretty much the final word in Greek voting for this referendum.  It's the map by the government. 

But then I noticed that ALL of the districts voted NO... except for that one little peninsula in the center of Northern Greece.  What's up with that? Did anyone else notice this one region that seemed different?  Here's a closeup (such as it is):  

When I looked at the Ministry's data, I couldn't figure out which district that was.  Odd.  

I also found the Guardian newspaper's interactive voting map, and it also didn't register any votes.  

So... What's going on?  

My curiosity was piqued.  So I flew to Google Maps, zoomed in on that part of Greece and was surprised to see a dashed line marking off this peninsula.  As you can see, the dropped pin associates the name "Mt. Athos" with this region.  

So... What is Mt. Athos?  A quick search shows that it's a UNESCO designated World Heritage site, and is described as "An Orthodox spiritual centre since 1054, Mount Athos has enjoyed an autonomous statute since Byzantine times. The 'Holy Mountain', which is forbidden to women and children, is also a recognized artistic site. The layout of the monasteries (about 20 of which are presently inhabited by some 1,400 monks) had an influence as far afield as Russia, and its school of painting influenced the history of Orthodox art." 

A bit more poking around and you can learn that the monasteries are reachable only by boat, and that only adult men can visit.  It's a serious, serious monastic existence. 

And they don't vote.  

Search Lessons

1.  When searching for data from another country, a site: restriction can often save you a LOT of time.  In this case, I got kind of lucky--the official Ministry of the Interior showed up in the second position.  

2.  Still, when looking for data, check around for alternative stories.  In this case, every site agreed with the official data (but that's not always true!).  Still, always check for consistency in the data. 

3.  Be sure you understand the bias of the sites.  You saw that proudly describes itself as leftist.  Not all sites do; sometimes you have to look at their written record to see where they're coming from (and take heed--sometimes a single source will have multiple writers with a variety of viewpoints, some right, some left, some from extreme positions--one writer doesn't (usually) define a news site's stance.  

4.  Sometimes clicking around on a site can save you a BUNCH of time and trouble.  Just as I was about the scrape the data from the Ministry's website, I found their charts, which were exactly what I was seeking.  Be sure you're not duplicating effort.  A minute of clicking can save hours of your labor! 

Thanks to all for everyone's hard work on this.  And thanks again to Miguel for the idea.  


Search on.