The Best Total Lunar Eclipses Visible From the US Between 1990-2050

One thing that was notable about the lunar eclipse on Sept 27-28, 2015 (and garnered substantial press at the time) was that the closest approach (perigee) of the Moon to the Earth actually occurred during the eclipse. This coincidence seemed like it would be quite rare, so I decided to look into it a bit further. Here are a few questions and answers that relate to lunar eclipses in general and perigee lunar eclipses in particular, with the intent to point out any that are unusual or special in some way:

What is a lunar eclipse?
A lunar eclipse occurs when the Moon travels through the Earth's shadow.

Are there different types of lunar eclipses?
Yes. The Earth's shadow has an 'umbra' and a 'penumbra'. If you are in the umbra and look towards the Sun, the Earth blocks the entire Sun. If you are in the penumbra you would see the Earth blocking only part of the Sun. Hence, there are three types of lunar eclipses: (1) Total eclipses where the entire Moon is in the umbra. During these, the only sunlight that reaches the Moon comes from the glow of the light refracted and scattered from the Earth's atmosphere, which would look like a fiery red ring everywhaere on the Moon. For this reason the Moon looks reddish during a total eclipse. Most total eclipses last about an hour or so. (2) Partial eclipses occur when only part of the Moon is in the umbra. The Moon looks like it has a chunk taken out of it. Partial phases also bracket every total eclipse, each lasting about an hour, typically. (3) Penumbral eclipses occur when the Moon misses the umbra entirely. From the Earth these are subtle to detect, with only a weak dimming on one side of the full Moon. If you were on that part of the Moon, the Earth would only block part of the Sun. Total eclipses are by far the most interesting to observe.

Here is a nice graphic of the September 2015 eclipse showing the Earth's umbra, penumbra, and the path of the Moon (from the Espenak NASA website, which has similar graphs for all the eclipses).

How frequent are all lunar eclipses and total eclipses?
We usually get two or three, including all types, each year. On average we get about one total lunar eclipse every year, but sometimes we'll get two in a year and sometimes none. Since total eclipses are only visible from the side of the Earth that happens to be facing the Moon at the time, any given location on the Earth will be able to see a total lunar eclipse about once every 2-3 years. The next total lunar eclipse visible from the US after 2015 occurs right around dawn on Jan 31, 2018. The next one in the middle of the night on Jan 20/21, 2019.

Why doesn't this occur every month as the Moon orbits the Earth?
The Moon's orbit is tilted about 5 degrees relative to the Earth's orbit around the Sun. So usually the Moon either misses the shadow above or below as it orbits the Earth. To line up the Sun, Earth and Moon well enough to get the Moon into the Earth's shadow you have to have the Moon lie nearly in the plane of the Earth's orbit around the Sun when the Moon is full. The points where the Moon crosses the plane of the Earth's orbit are called nodes. Lunar nodes are aligned with the Sun and Earth every 6 months, and during that time we get eclipses. In 2015, the eclipse seasons are April and September.

Do eclipse seasons change?
Yes. In 2016 the lunar and solar eclipses occurred in March and September, and in 2017 they shift to February and August. Eclipses occur a little earlier each year, and return to their original starting point after about 18.6 years. This `regression of the nodes' is caused by the lunar orbit wobbling like a top around the Earth in response to the torques acting on it by the bulge of the Earth and the pull of the Sun, both of which are inclined to the lunar orbital plane.

What is a blood Moon?
No astronomer I know uses this term, which sounds more at home in a vampire novel. Usually it simply refers to an eclipsed Moon, which appears red because sunlight from all the sunrises and sunsets on Earth are simultaneously visible on the lunar surface. A few people have used the term to refer to a tetrad, where four successive lunar eclipses are all total ones, but this usage makes no sense to me because whether or not penumbral or partial eclipses are interspersed between the total ones has no bearing upon the color of the Moon during an eclipse.

What is a SUPERMOON?
Another relatively recent item of popular culture. This one refers to a full Moon that occurs when the Moon is near perigee, it's closest approach to the Earth. There does not seem to be any criteria as to how close the Moon needs to be to qualify as a `supermoon' (see my article on supermoons).

Does the fact that perigee occurs during eclipse make an eclipse special?
It does makes it very unusual. After the September 2015 event, the next time perigee occurs during an eclipse will be on Oct 8, 2033 (one saros period in the future). In terms of how the eclipse appears though, a supermoon eclipse is not going to look a whole lot different from a typical eclipse because the amount of the increased apparent size of the Moon at perigee is rather minimal. In addition, because of variations in the lunar orbit, and your location on the Earth, the Moon can get closer than it is on 9/27/15 even when it is not at perigee (see below). Really the best part about the 9/27/2015 eclipse was that it occurred at a very convenient time that made for pleasant viewing for most people.

So which total eclipses are the best ones in the next several decades?

It seems to me for an eclipse to be a good one for US viewers it needs to be (in order of preference)

  • Visible from the US
  • Total (eliminate partials and penumbrals)
  • Central eclipse occurs at a convenient time of the evening
  • Have a long total phase
  • Have the Moon be close so it looks bigger (if we can get it, why not?)

    In the graphs below I have sorted through all the total lunar eclipses visible from the US (defined as middle eclipse being when the Moon is above the horizon for viewers in the middle of the country) between 1990 and 2060. The dates listed for eclipses are UT, and therefore refer to the date after midnight for the night in question for the US. That is, the eclipse at 9pm on the evening of Sept 27, 2015 is labeled as '9/28/15', while the eclipse at 3 am on August 18, 2054 is labeled as '8/18/54'. Red triangles indicate eclipses where the Moon's apparent radius is at least 5% larger than normal, so its area appears at least 10% larger than normal. Blue circles indicate all other cases. Here I have defined normal to be 379000 km, equal to the semimajor axis of the Moon's orbit minus about 4000 km to compensate for a typical position on the Earth that sees the full Moon.

    The graphs show that 10 out of the 34 total lunar eclipses visible from the US in this 70-year interval have an unusually large Moon, with 6 of these visible in the evening hours before midnight. The eclipse of 9/27-28/15 is one of these. The graph below shows that only the eclipse of 1/20-21/19, also well-placed in the evening sky from the US, appears to have a slightly larger diameter for the Moon during this time interval.

    Interestingly, even though perigee occurs during the 9/27-28/15 eclipse, and perigee occurs 14 hours after the eclipse on 1/20-21/19, the eclipse on 1/20-21/19 is the closer of the two. How can this be? The main factor here is that the 1/20-21/19 eclipse occurs around midnight for US observers. At midnight the full Moon is highest in the sky, and the observer is closest to the Moon than during other hours of the night. The 9/27/15 eclipse occurred in the early evening for US observers and so was at a slightly larger distance than if the eclipse were to occur at mignight (the Earth's radius is about 6400 km, about 2% of the distance to the Moon, so the effect of different locations on the lunar distance is typically about 1%).

    However, there is another more interesting effect on the distances that has to do with the lunar orbit. The eccentricity of the Moon's orbit has many periodic variations, and gets more eccentric when the major axis of the Moon's elliptical orbit aligns with the Sun. Have a look at the instantaneous eccentricity (part of what is known as the osculating elements of the orbit) of the lunar orbit from Figure 4.6, again from the excellent NASA site. The closest approaches of the Moon to the Earth occur when the eccentricity is highest and the orbit is more elliptical. Even so, the eccentricity of the lunar orbit is never higher than about 0.078, so the maximum variation in the apparent lunar diameter is never more than about 15%, and can be as low as 5% at times when the orbit is most circular. This difference in size is noticeable, but not dramatic. According to this lunar calculator, the 1/20-21/19 perigee is actually slightly farther away (357344 km) than the 9/27-28/15 perigee (356876 km). There are slightly closer perigees, for example on Nov 14 2016 (356511 km; angular size 34.1 arcminutes when viewed from the optimal location on the Earth, in this case near Hawaii), but no eclipse occurs during that full Moon because the lunar nodes do not align with the Sun during November of that year. All this is discussed in my supermoon article.

    In celestial mechanics one can describe the lunar motion in terms of a stable orbit that undergoes a perturbation by a third body (either the Sun or the bulge of the Earth). The equations of motion are then reformulated in terms of a perturbation function. The orbit is described by 6 parameters, two for shape (a = semi-major axis, and e = eccentricity), three for the orientation of the ellipse in space (i = inclination, node = location of node, and omega = location of perigee), and one to place the object on the orbit (tau = time of closest approach). One can solve the perturbation equations for each orbital parameter and determine secular, short-term-periodic, and long-term-periodic components to the equations. The NASA graph shows that these perturbations generate periodic changes in the eccentricity, with rather intricate variations with time. The Moon's orbit is quite tricky to get right!

    Fun fact: The eclipse on Jan 1, 2048 is actually ongoing at midnight. So on that date we can ring in the new year with a total eclipse of the Moon!


    Back to Hartigan's Home Page
    Patrick Hartigan
    hartigan@sparky.rice.edu