The Science Behind Your Star

Every star we identify comes from a documented calculation you can check yourself. This page explains, in plain language, how the sky of one moment is reconstructed — the coordinates, the time conversion, the star catalog and the selection rule — including the parts we deliberately do not model.

A Star Named is an astronomical memorial service that reconstructs the sky at the recorded time and location of a person’s or pet’s passing and identifies the star closest to the zenith.

1. Location — where the sky is measured from

The place you enter is resolved to a latitude and a longitude. Latitude fixes which part of the sky passes overhead: at any instant the zenith’s declination equals the observer’s latitude. Longitude shifts local time and the sky’s rotation angle — it decides when that point comes overhead (section 3).

Elevation is not used — and here is why

We do not adjust for height above sea level. Altitude changes sunrise and sunset by a minute or two, but the star we identify lies within a few degrees of the zenith, where elevation has practically no effect. It also means we never ask you for a number you may not know. If a moment falls exactly at dawn or dusk, elevation could in principle shift the pick — we accept that small uncertainty instead of inventing precision.

2. Time — from a clock reading to an instant in UTC

You enter the local civil time at that place — the time that was on the clock. We convert it to UTC using the IANA timezone database, which carries the real history of clock rules for the region, including daylight saving. Every astronomical computation then runs on UTC alone.

3. The astronomical calculation — sidereal time, hour angle, altitude, azimuth

Given an instant in UTC, a latitude and a longitude, we find the point of the sky directly overhead:

4. The star catalog — HYG v4.1

Candidates come from the HYG catalog, a real public catalog merging the Hipparcos, Gliese and Yale surveys: J2000 positions (RA/Dec), distances from parallax, brightness and spectral type. We include 107,838 stars covering all 88 IAU constellations; every record is traceable to its source IDs. The full accounting lives on our data page.

The catalog spans magnitude 0.03 to magnitude 12 — far below the naked-eye limit of about 6.5. The star nearest the zenith is often a faint one. That is the honest trade-off of choosing by position rather than brightness, and your certificate states the magnitude so you know what to expect.

One more honest note: positions are stored at epoch J2000, and we do not yet propagate each star’s proper motion to the date of observation. For nearly every star that amounts to a few arc-seconds over a human lifetime — far below the selection resolution — but for a handful of very fast nearby stars it can matter. The correction code is in place; it will be switched on as proper-motion values are added to the catalog rows.

5. Choosing the star — the exact rule

Stated the way it is implemented:

6. Check it yourself

You do not have to take any of this on faith — Stellarium (free and open source) or any planetarium app can confirm the result:

7. Known limitations

Go deeper


星背后的科学

我们找出的每一颗星,都来自一套你可以亲手复核的计算。这一页用平实的语言讲清楚:某一刻的星空是如何被重建的——坐标、时间换算、星表与选星规则——包括我们刻意没有建模的部分。

A Star Named 是一项天文纪念服务:它重建一个人或一只宠物离世时刻、在其所在地的星空,并找出最靠近天顶的那一颗真实恒星。

1. 位置:从哪里量这片天空

你填写的地点会被解析为一组经纬度。纬度决定天空的哪一部分经过头顶:任意时刻,天顶的赤纬都等于观测者纬度。经度决定时刻与天空转过的角度——也就是那个点何时升到头顶(见第 3 节)。

高程:不使用,以及为什么

我们不按海拔高度做修正。海拔会让日出日落时刻差上一两分钟,但我们找出的星位于天顶附近几个角距之内,在那里海拔几乎没有影响。这也意味着我们不会向你索要一个你可能并不知道的数字。如果某一刻恰好落在黎明或黄昏,海拔理论上可能影响选星结果——我们宁愿接受这点不确定性,也不虚构不存在的精度。

2. 时间:从钟面读数到 UTC 时刻

你输入的是当地民用时间——当时钟上显示的时刻。我们用 IANA 时区数据库把它换算为 UTC;该数据库记录了各地区真实的时钟规则历史,包括夏令时。此后所有天文计算只依赖 UTC,不再做任何时间假设。

3. 天文计算:恒星时、时角、高度角与方位角

给定 UTC 时刻、纬度与经度,我们求出天空正头顶的那一点:

4. 星表:HYG v4.1

候选星来自 HYG 星表——一张真实的公开星表,合并了依巴谷(Hipparcos)、格利泽(Gliese)与耶鲁亮星目录:J2000 位置(赤经/赤纬)、由视差换算的距离、亮度与光谱型。本站收录其中 107,838 颗,覆盖全部 88 个 IAU 星座,每条记录都可追溯到来源编号。完整的统计与版本见我们的数据页。

星表的星等范围是 0.03 等到 12 等——远暗于肉眼极限(约 6.5 等)。最靠近天顶的那颗星,常常是一颗暗星。这是"按位置而非按亮度选星"的诚实代价;证书上会写明星等,让你心里有数。

还有一条诚实说明:坐标以 J2000 历元存储,我们目前尚未把每颗星的自行传播到观测日期。对绝大多数恒星,几十年也只有几角秒——远低于选星的角距分辨率——但对极少数高速近邻星会有影响。修正代码已经就位,待自行数据补入星表行后即会启用。

5. 选星:与实现一致的精确规则

规则按实现原样陈述:

6. 亲自复核

这些都不需要你凭信任接受——用 Stellarium(免费开源)或任何星空软件即可核对:

7. 已知限制

继续深入