True Solar Time in Bazi

A birth certificate records civil time: the time shown by clocks under the time-zone and daylight-saving rules in force at that place and date. A Bazi calculator may instead derive the Hour Pillar from local apparent solar time — what a properly constructed and calibrated sundial would indicate at that location and instant. Converting the recorded clock time to that solar reading requires the historical UTC offset, birthplace longitude, and the sun's seasonal variation. Civil and solar time are not always the same minute, or even the same two-hour branch.

That difference is one reason two calculators can return different Hour Pillars from the same recorded birth details. But it is not the only reason: calculators may also use different historical time-zone data, birthplace coordinates, solar-time formulas, or Bazi day-boundary conventions. This page documents the method used by BaziCalculator.ai so the result can be checked rather than taken on trust.

Three Kinds of Time You Should Not Mix Up

A time-zone label alone is not enough. For example, modern China uses UTC+8 across a wide longitude range, and the civil-time rules in 1990 were not the same as they are today. The calculation therefore needs the birthplace, local date, and local recorded time together.

The Formula Used by This Calculator

This is the sign convention in NOAA's General Solar Position Calculations. It is useful because it accepts the actual UTC offset directly: when a birthplace was observing daylight saving time, there is no need to subtract a second, hand-written “DST hour” later. Doing both would double-count the clock change.

The corrected timestamp is also the input to this site's Da Yun starting-age calculation. That guide shows how historical daylight saving time can change the birth-to-Jie interval and exposes the exact boundary in a calculation receipt.

Step 1: Resolve the Historical UTC Offset

Time zones are political histories, not permanent longitude bands. Governments change offsets, start or stop daylight saving time, and occasionally use unusual increments. The IANA Time Zone Database records these changes for representative locations and is periodically updated when rules change.

The production engine uses an IANA zone name such as America/Chicago or Asia/Shanghai and asks the runtime for the offset that applied to the supplied local date and time. This catches details that a fixed “UTC+8” or “subtract one hour in summer” rule misses. In the regression suite, China is UTC+9 on 15 June 1990 because daylight saving time was in force, while Lord Howe Island supplies a 30-minute — not 60-minute — DST increment in January 2025.

Whether a birth date falls inside a daylight-saving era is one of the first things to check, because one hour is half a branch window. The table distinguishes the rules encoded in the IANA data queried by the engine from archival history: those are not always identical, especially before 1970. Hong Kong's official record comes from the Hong Kong Observatory; Taiwan's 1969 exception is documented by Taiwan Historica but is absent from the current Asia/Taipei rules.

Step 2: Correct for Longitude

Relative to the mean sun, 360 degrees corresponds to 24 hours of mean solar time, so one degree of longitude corresponds to exactly four minutes on that scale. This is distinct from Earth's roughly 23-hour-56-minute rotation relative to the stars. A civil time zone assigns one clock to a region, but solar noon still moves continuously with longitude. A city west of its zone's nominal central meridian reaches local solar noon later by the civil clock; a city east of it reaches solar noon earlier.

Step 3: Add the Equation of Time

Even after longitude is accounted for, apparent solar time does not run at a perfectly uniform rate through the year. Earth's orbital eccentricity and axial tilt cause the apparent sun to cross the local meridian at uneven clock intervals. The Equation of Time (EoT) describes the difference between apparent solar time and mean solar time.

In the NOAA fractional-year approximation used here, a positive value means apparent solar time is ahead of mean solar time; a negative value means it is behind. The curve is roughly −14 to +16 minutes over a representative year. That term is smaller than the longitude correction for many cities, but it can still decide a chart that lies close to a two-hour boundary. The approximation's own error must therefore be included in the boundary uncertainty, not hidden by minute-level rounding.

Swipe to inspect the diagram · Tap to open full size

Worked Example: Chicago, 15 June 1990

Suppose a birth record says 12:00 on 15 June 1990 in Chicago, using longitude 87.63°W and the IANA zone America/Chicago. The current production engine returns the following intermediate values:

1. The historical offset at that local time is UTC−5. Chicago was observing daylight saving time, 60 minutes ahead of its standard offset.
2. Relative to the UTC−6 standard meridian, the longitude component is +9.48 minutes.
3. The NOAA Equation of Time for that date and hour is −0.03 minutes in the engine's rounded output.
4. The direct NOAA formula gives a total correction of −50.55 minutes.
5. Recorded 12:00 therefore becomes 11:09 true solar time after normalization. Under this site's branch windows, that remains the Wu hour (11:00–13:00), but it is only nine minutes from the preceding boundary.

When the Correction Changes the Hour Pillar

The same recorded noon does not produce the same corrected branch everywhere. On 15 June 1990, the regression examples for Beijing and Chengdu both resolve to UTC+9 because China was observing daylight saving time. Their more westerly longitudes then pull local apparent solar time back far enough to cross from the recorded noon's Wu window into the preceding Si window.

The Two-Hour Windows Used on This Site

A correction matters most near one of these boundaries. If the recorded time is rounded, reconstructed from memory, or written during a repeated DST hour, the mathematical output can be precise while the input remains uncertain. The responsible response is to show that uncertainty — not to present the final branch as more certain than the record allows.

Why Two Bazi Calculators May Still Disagree

• One calculator uses recorded civil time directly while another applies local apparent solar time.
• One uses the correct historical UTC offset while another applies today's offset or a fixed one-hour DST rule.
• The city coordinates differ, or one tool substitutes a regional center for the actual birthplace.
• The Equation of Time formula, sign convention, or leap-year handling differs.
• The tools use different Zi-hour or day-boundary conventions near 23:00 and midnight.
• One tool cuts the Bazi Year Pillar at the Lichun solar term in early February while another reports the popular zodiac from Lunar New Year or uses January 1st. A January 1st boundary can disagree with Lichun for every birth from January 1st until Lichun; a Lunar New Year boundary disagrees during the interval between Lunar New Year and Lichun, whichever comes first (this site labels both legitimate conventions).
• One tool rounds intermediate values before determining the boundary while another preserves full precision.

Limits and Calculation Conventions

• Birth record: a remembered or rounded time can dominate the uncertainty, especially within about 30 minutes of a branch boundary.
• Coordinates: a city-center longitude is appropriate for most users, but it is not the exact delivery-room coordinate. One degree equals about four minutes.
• Historical civil time: IANA notes that pre-1970 local-time history is harder to model comprehensively and may be incomplete for some locations.
• Pre-1949 China: the Central Observatory proposed five zones in 1918, from Kunlun (UTC+5:30) through Longshu (UTC+7) to Changbai (UTC+8:30). The plan initially caught on mainly in UTC+8 coastal areas, was displaced in theory by UTC+7 during the war, and was officially restored with modifications in March 1948; surviving evidence does not establish uniform observance. This calculator does not reconstruct that history. Most Chinese cities use Asia/Shanghai, which supplies UTC+8 and Shanghai's wartime DST rules, while Xinjiang cities use Asia/Urumqi, an approximate UTC+6 model from 1928. Depending on the place and clock, a five-zone interpretation can differ from a Shanghai-based result by 30, 60, 120, or 150 minutes. Identify the clock represented by the record before trusting a near-boundary Hour or Day Pillar.
• Astronomical model: the 1990 daily comparison above found up to about 0.76 minutes between NOAA's fractional-year approximation and USNO ephemeris data. The model is suitable for this calculator's minute-level correction, but a result that close to a branch boundary needs a higher-precision check.
• DST transition: repeated or skipped local times require interpreting the original record, not just applying arithmetic.
• Bazi day boundary: the public calculator's current sect 2 default starts the Zi Hour branch at 23:00 but rolls the Day Pillar at civil midnight. The engine can also calculate sect 1, where 23:00–23:59 belongs to the next Day Pillar; the public form does not currently expose this as a switch.

Regression Matrix

The current test suite imports the production correction function instead of reimplementing the math in a separate test. Each case checks the corrected minute, the resulting branch, and the historical DST increment. The expected display values are: