One Doorstep, Fifty-Two Sunsets: The Horizon Diary as a Year-Long Practice
Stand in the same spot once a week and write down two numbers: the clock time of sunset and the compass bearing of the point on the horizon where the Sun disappears. That is the whole practice. One fixed viewpoint, fifty-two entries, a year of geometry passing through a doorway.
The diary is not an instrument. It will not tell you your latitude to a tenth of a degree. What it does is make the seasonal swing of the sunset visible as a line of ink on a page, and it gives you a reason to look west on a schedule rather than by accident.
Why bearing and time, and not just one or the other
Azimuth and clock time drift for different reasons, so recording both makes the diary more useful than recording either alone.
Azimuth is the angle measured clockwise from true north to the point on the horizon directly below the Sun. The U.S. Naval Observatory’s solar position service computes it that way, and its notes are explicit that the values are referenced to true north, not magnetic north (USNO Altitude and Azimuth). That distinction matters the moment you bring a compass to the doorstep. A hand compass points at the magnetic pole, and the gap between magnetic and true north varies by location. The USNO page points readers to NOAA’s geomagnetic resources for declination rather than assuming a value.
Clock time drifts for a different set of reasons: the equation of time, the tilt of your time zone’s meridian relative to your longitude, and daylight saving. The USNO’s year-long sunrise and sunset tables are issued in local standard time on a 24-hour clock, so a table entry of 1836 means 6:36 p.m. standard time (USNO One-Year Table). If your clock is on daylight saving, the table and your wristwatch will disagree by an hour for part of the year. Write down which clock you used. Future you will want to know.
There is a third reason the two numbers belong together. The observed moment of sunset is not the geometric moment the Sun’s center crosses the horizon. Refraction lifts the apparent Sun, and the USNO notes that its altitude values include standard atmospheric refraction when the object is above the horizon. The same page warns that observed rise and set times are affected by random changes in local atmospheric conditions that cannot be accurately modeled, which is why USNO tabulates rise and set to the nearest minute only (USNO Rise/Set/Transit). Your entry is a real observation under real sky; the table is a model. The mismatch between them is information, not error.
Setting up the diary
Choose a marker you will not move: a doorframe, a gatepost, the edge of a window. The point is to hold latitude, longitude, and horizon fixed so that any change you record is seasonal rather than positional. A doorstep is a legitimate instrument precisely because it does not move.
You need a true-north reference before your first bearing entry. The USNO azimuth convention is clockwise from true north, so a bearing of 270 degrees is due west and 240 degrees is west-southwest. If you use a magnetic compass, apply the local declination correction, or use a phone compass app that reports true bearings. Do not guess the correction; look it up for your location.
A simple entry has five fields:
- Date and day of the week.
- Clock time of the Sun’s disappearance, and whether the clock is standard or daylight time.
- Bearing of the setting point, in degrees clockwise from true north.
- Marker note: which fixed object on the horizon the Sun set behind, in body-scale terms. A fist at arm’s length spans about 10 degrees; four upright fingers against the horizon correspond roughly to an hour of remaining daylight. These are rough anchors, not measurements.
- Sky note: one line on cloud, haze, or clarity. This is what lets you interpret a gap in the record later.
If you miss a week, write the next entry and note the gap. A diary with holes is still a diary. The year does not require perfect attendance.
Using the official tables as a check, not a substitute
The USNO’s one-day altitude and azimuth service will give you the computed azimuth of the Sun at any time on any date between 1700 and 2100 (USNO Altitude and Azimuth). The one-year table will give you sunrise and sunset times for a full year at your coordinates (USNO One-Year Table). The rise/set/transit service adds azimuth at rise and set for the Sun and other bodies (USNO Rise/Set/Transit).
Use them after you have written your entry, not before. If your bearing is consistently off by the same amount, your true-north reference is probably off. If your time is consistently early or late by a fixed amount, check whether you and the table are using the same clock convention. If the mismatch wanders from week to week, that is the sky doing what the USNO says it does: local conditions that no table can model.
NOAA’s solar calculator is a second option, and it states the azimuth convention plainly: measured clockwise from true north to the point on the horizon directly below the object (NOAA Solar Calculator). The same page carries a notice that the calculator is no longer actively supported or maintained, and it cautions that actual observed values may differ from its results because of variable atmospheric conditions and algorithm uncertainty. That caution is not a reason to avoid it; it is a reason to treat any calculator as a cross-check on your own record rather than as the record itself.
Reading the year in two curves
After a few months, plot two lines: bearing against date, and clock time against date. They will not move together. The bearing line shows the Sun’s setting point sliding along the horizon and then turning back. The time line shows the evening lengthening and shortening, with a shape set by your latitude and your time zone’s offset from your longitude.
The bearing spread is the more striking of the two. Sunrise azimuth is not 90 degrees except near the equinoxes, and the spread between the June and December bearings widens as you move away from the equator. A table computed from the NOAA general solar position equations gives, for example, a June sunrise azimuth of about 63.5 degrees at 25 degrees north and about 44.4 degrees at 55 degrees north, with the December values mirroring on the other side of east (Calculator Academy solar azimuth reference). That page describes itself as using NOAA’s equations; it is not an official NOAA publication, so treat its numbers as illustrative of the method rather than as authority. The point stands: the seasonal swing is small near the equator and large toward the poles.
Your own diary will show the same geometry without any table at all. The first time the setting point reverses direction along the horizon, you have found a solstice without naming it.
The southern hemisphere and equatorial version
In the southern hemisphere the same seasonal bearing numbers apply with the seasons swapped, and the noon bearing becomes north rather than south. The Sun still rises in the east and sets in the west on the equinoxes, but its arc leans the other way, and the solstice that gives the longest evening is December rather than June. A diary kept in Sydney or Santiago will show the setting point reaching its southern extreme in December and its northern extreme in June, the mirror of a northern-hemisphere record.
Near the equator the swing compresses. The Sun rises and sets close to due east and due west all year, and the bearing line on your plot will look almost flat compared with a mid-latitude record. The time line, by contrast, stays comparatively even. An equatorial diary is still worth keeping, but its lesson is different: it shows how little the horizon moves when you are close to the plane of the Sun’s annual path.
What the diary is not
It is not a way to determine your latitude, longitude, or the date to a stated precision. It is not a substitute for the USNO tables if you need a computed value. It is not a record of the Sun’s geometric position, because refraction and local horizon height both intervene. And it is not a claim about who first kept such a record; the practice is old enough that its origins are not something this page can settle.
What it is: a year of looking from one place, with two numbers and a line of sky note per week. The geometry of the year passes through your doorway whether or not you write it down. The diary is a way of noticing that it did.
Questions readers ask
Do I need a compass?
You need a true-north reference. A compass with a declination correction works; a phone app that reports true bearings works; a known landmark with a published bearing works. The USNO azimuth values are referenced to true north, so a magnetic bearing without correction will not match them.
What if I miss weeks?
Write the next entry and note the gap. The diary is a practice, not an audit. A year with thirty entries still shows the seasonal turn.
Why does my time not match the USNO table?
Check three things first: whether the table is in local standard time while your clock is on daylight saving; whether your longitude and time zone offset are entered correctly; and whether you recorded the moment the upper limb disappeared or the moment the center crossed. The USNO tabulates to the nearest minute because local conditions cannot be modeled more precisely.
Does the southern-hemisphere version need a different table?
No. The USNO services accept southern latitudes directly. The seasons swap in the output, and the noon bearing becomes north rather than south. The method is identical.
Is the NOAA calculator still reliable?
NOAA states that the calculator is no longer actively supported or maintained and that observed values may differ from its results. Use it as a cross-check, and prefer the USNO services when you need a documented, maintained table.