Wednesday, September 22, 2021

SearchResearch Challenge (9/22/21): Time and date from a shadow?


Photos usually have more information in them than you might expect... 

This is a photo of the SkyTree in Tokyo, Japan.  (That is, 東京スカイツリー, Tōkyō Sukaitsurī) 

It's the tallest tower in the world--taller than the Canton Tower (which was loftiest until 2011) and is (currently) the second tallest structure in the world after the Burj Khalifa (which is 829.8 m or 2,722 ft). It's a neofuturist design that captures the genius loci of modern Tokyo.  

Comparison of tallest towers. By timesdad, P/C Wikimedia.

Here's another photo of the SkyTree, this time taken from the Bing Maps image.

(I used Bing Maps satellite imagery because they had a beautifully clear pic with a very crisp shadow of the tower.)  

As you also know, a shadow can reveal much--in particular, it can tell us everything about when the photo was taken.  

Here's another photo of the tower shadow, with a red arrow showing the length of the shadow.  It goes from the center of the tower's base to just beyond the bridge over the Sumida river.  

Our Challenge for today is simple: 

1.  What day and time-of-day was this photograph taken?  

This might seem crazy-hard, but it's actually fairly straight-forward.  (And no, you don't have to remember your elementary trig to figure it out--remember that you have search engines to help you out!)  

I wanted to pose this Challenge because it's another example of "if you know something is possible, you can figure out how to do it with a bit of search."  

This really IS straight-forward--just don't get scared along the way.  

Be SURE to let us know how you figured the answer; we all want to learn along with you! 

Search On!  


  1. I do not have a solution but I have learned what I need to know before finding the time of day: the day of year (or does day include year also?) which I should be able to get from the elevation and azimuth of the Sun. This makes sense because I am acquainted with a sundial that has a correction factor for time of year.

    I used [latitude longitude sky tree tower] and got Latitude and longitude coordinates are: 35.710064, 139.810699.

    Looking at it on Google Maps I might could estimate the length and direction of the shadow, which would give the position of the sun, but that would be only an estimate. There must be some search engine which can measure that from the photo of the shadow especially given Dan’s red line – maybe? This search engine would have to know the exact location of the other Tokyo landmarks shown relative to the tower.

    Details of how I got here will follow.

  2. The beginning of the story:

    I started with the obvious:
    [time and date photo was taken] which was about retrieving a photo’s date from a device. Dead end here.

    So I changed it:
    [time and date photo was taken by shadow]
    The example here was on the Israel National Trail. The month was known and weather data was used to pinpoint the day. Interesting, but not relevant here. According to the article, “Now that we have the day and the location, finding the exact time is the easiest part.”

    Other sources were about estimating the date or finding time of day given date.
    So I got more specific
    [find day of year photo was taken using shadow]

    This gave me similar results as previous queries, plus
    This article pointed out that we do not need to know the actual length of the object, but only the proportional length of the shadow compared to the object. True, but not helpful.
    There is a tool called SunCalc but the location and date must be known so this is not useful. Why am I not getting results about the day of the year?

    I learned about something called chronolocation, which is not helpful because I know the location. What is the inverse of this?

    I tried
    [find length of a shadow from a photo] This will tell me the length a shadow will be given date and location. Again, I can see the shadow but need the date.

    Maybe I can use a clue from the challenge.
    [distance Tokyo sky tree to just beyond the bridge over the Sumida river] This gave walking/driving distances and information about visiting the tower.

    So I tried [distance as the crow flies Tokyo sky tree to the Sumida river] and got distances to other cities from Tokyo and information about the river.

    I am getting the same results. I think this might be a language issue and need some rewording of the queries.
    I am going to try [sundial corrects for time of year] and [what can a shadow tell about day and time].

  3. The math teacher in me really wanted me to to do some calculations, but I heeded Dan's warning and skipped that (for now).

    Using a ruler function on Google Earth Pro (or equivalent), I was able to measure the shadow length to be approximately 880.1 meters with a heading of 158.7° degrees. With that, I needed a coordinate location (readily available from Google Earth/Maps and Wikipedia), the height of the tower (also available in Wikipedia) and a good tool that does the calculations for us.

    Some results from my "sun tracker for any place on earth" and "determine sun position by date and time" Google Search gave me a few results; the most useful of which I found to be SunCalc.

    Equipped with this information, I
    • centered the bullseye on the Skytree tower
    • entered 634 for the height in meters
    • selected February 14 as my initial date to test (noting the sun had to be low in the sky to produce such a shadow)
    • adjusted the date and time until the calculated shadow length and azimuth matched the ones I measured.

    Conclusion: This photo was taken on February 06 at 10:40 UTC+9, when the shadow length was 880.7 meters and an azimuth of 157.6. This of course fails to take into account variations from year to year.,139.8107,16/2021.02.06/10:26/634/3

    1. By all means, go ahead and do the math!! (I just wanted everyone to know that there are simpler workarounds.)

    2. If I am able to find the time, I shall. But with a such a dynamic system (accounting for the analemma in addition to varying angle as the sun traverses horizontally and vertically throughout the day), it may be a while.

      In the meantime, I shared this on Facebook, and a college friend reminded me that unless it was taken on a solstice, there should actually be TWO dates that produce that particular shadow.

      So, a little math: my initial date was 47 days after the Winter Solstice. Applying a principle of symmetry, the other date should be 47 days before the Winter Solstice: 04 November. With that, I set the SunCalc to 04 Nov, adjusted the time a little bit, and sure enough at 10:07 UTC+9, the shadow length was exactly 880.8 meters with an azimuth of 156.9°!

      So which is the correct answer? Eh, flip a coin OR research weather data for those particular dates, note which one had more sunny days than the other (assuming there is in fact a seasonal difference), and then make a statistical guess that is the more likely of the two. I think I'm comfortable calling it good for now.

  4. I tried and think a lot about the How.

    I'm almost sure but just to confirm. You want us to find when the Bing photo was taken, right?

    I searched things like [date photos taken Zoom Earth] There was an easy way, just zooming but it was removed.

    Tried other queries without luck.

    Then I searched [ When was Google maps photos taken]

    2020:Find the Exact Date When a Google Maps Image was Taken

    I haven't been able to try as at the moment I don't have a desktop

    1. With [how to tell when a photograph was taken bing maps after:2019]

      2021 (apparently)

      Viewing Bing Maps imagery dates

      I tried on mobile. I couldn't. Maybe someone else can

  5. This comment has been removed by a blog administrator.

  6. Hi everyone.

    I first tried to get data directly from the image at Bingmaps, but could not find any.
    I found some old posts about getting the date from the Bingmaps API, so I did that. Only got the date 02-22-2021. I had to get a bingmaps API key.
    Then searching about getting the time from shadows (determine shadow by time of day) I got to this site:
    There I went to "Sun Position"
    There I just entered the coordinates and date and got a map with a sundial on it showing the original bing image.
    I could tell by the dial that the image was taken at 13:00

  7. Sharing a fun approach one of my students is trying (still awaiting to see if it works out). He noticed some boats on the river near the SkyTower Tree and decided to see if he could locate likely ports nearby, use their logs to identify the boats and then use the records to indicate when those boats would have been in their approximate respective locations.

    Though I think this approach will require more effort than he is likely willing to make (certainly more so than I am!), I think it is a pretty interesting and viable solution nonetheless.

  8. This comment has been removed by a blog administrator.

  9. This comment has been removed by a blog administrator.

  10. This comment has been removed by a blog administrator.

  11. This comment has been removed by a blog administrator.

  12. you must have the 23 blahs…