**About Lumavik — Who We Are and How We Work**

> Who builds the Lumavik aurora forecast, where the data comes from, how the site makes money, and the editorial rules we hold ourselves to.

Source: https://lumavik.org/about/

[Home](https://lumavik.org/) /About

About

# A small team, a public dataset, and a strong preference for saying no.

Lumavik is an aurora forecast app for iPhone and a set of free space-weather tools on
the web. It is built by a small independent software team, not by an observatory, a
university or a government agency. We think you should know that before you decide how
much weight to give anything here.

## Why this exists

The project started from an ordinary frustration. The data needed to forecast aurora is
free, excellent and public — NOAA's Space Weather Prediction Center publishes the
planetary K-index, the three-day geomagnetic outlook, the OVATION auroral-oval model
and real-time solar wind measurements, all in the public domain. And yet the experience
of trying to answer one simple question, "is tonight worth it where I am", was
consistently bad. The authoritative sources publish instrument output for people who
already understand it. The readable sources are attached to something being sold. And a
lot of the apps in between take a planetary average, put a friendly face on it, and
present it as a local forecast, which it is not.

We built the thing we wanted: one number for one location, computed against the
coordinate the aurora actually responds to, with the limiting factor named every time.

## Where the data comes from

All of it comes from the NOAA Space Weather Prediction Center. The Kp index and its
one-minute estimated version, the three-day forecast, the OVATION model output and the
L1 solar-wind measurements are theirs. We are not affiliated with NOAA, we do not
represent them, and we do not improve on their measurements. What we add is the
conversion from a planetary dataset to a local answer: geomagnetic coordinates, the
auroral-oval boundary, solar geometry for darkness, and an honest account of the
uncertainty in all of it.

Where we approximate, we say so on the page. Geomagnetic latitude uses the tilted
dipole rather than the full field model. Twilight uses a standard low-precision solar
position series. Both are good enough to plan a night around and both can be off by a
small amount, and we would rather tell you that than present four decimal places.

## How we make money, stated plainly

The app is free and shows no advertising. A Pro subscription adds tunable alerts,
several saved locations, the hourly and solar-wind detail, and the search for clearer
sky nearby. That is the entire business model. We did plan a banner for the free tier
once, and this site said so; we dropped it before launch, on the grounds that a screen
you check in the dark, on a night that may come to nothing, is a poor place to put
something blinking for a stranger. This website carries
no advertising and no affiliate links: we do not take commission on tours, hotels,
flights or camera gear, and if we ever did we would label it on the page rather than in
a footer nobody reads.

Being upfront about this matters here more than in most categories, because the
incentive to overstate a forecast is real and obvious. An app that says yes more often
gets opened more often. We have deliberately built the opposite, and the honest reason
is not virtue: an alert that drags somebody into a cold field for nothing is how you
lose them permanently.

## Editorial rules

Everything published on this site is written under a small set of rules that we would
rather be held to publicly than keep to ourselves.

We do not invent numbers. If a figure would need a study to support it and we do not
have one, we describe the mechanism qualitatively instead of manufacturing a
percentage. Forecast reliability is discussed in terms of lead time and physics, never
as a made-up accuracy rate.

We do not promise sightings. Aurora visibility depends on four independent conditions
and the weakest one decides your night, and one of them — cloud cover — is outside
every geomagnetic model in existence. Any page here that discusses odds says so.

Our comparison pages name situations where a competitor is the better choice, and mean
them. Facts about other apps are taken from Apple's public listing data on a stated
date and from the developers' own store descriptions; we do not speculate about
features we cannot verify, and we do not accuse anyone of anything. If we have got
another team's product wrong, we want to hear about it and we will correct the page.

We correct errors visibly. Nothing here is edited silently after somebody points out a
mistake. If you find one, write to
[support@lumavik.org](mailto:support@lumavik.org).

## How the forecast is actually computed

There is no machine learning in this app and no proprietary model, and we would rather
say that than let the word "algorithm" imply something it does not. The chain is short
and each step is checkable.

We take the current estimated Kp and NOAA's forecast blocks. We convert your
coordinates to geomagnetic latitude, because that is the coordinate the auroral oval is
organised around and it can differ from your map latitude by several degrees. We
compare that against the approximate equatorward boundary of the oval at each Kp step,
which gives the minimum activity your location needs before the aurora reaches you at
all. We compute the solar elevation across the coming night to find out when, and
whether, your sky gets dark enough for a faint display to be visible. Then we report
the weakest link in that chain alongside the score, because the weakest link is what
actually decides your night.

Everything in that list runs on the device. The only thing that leaves it is a
coordinate rounded to half a degree, and only if you have asked for alerts, because a
notification has to be decided somewhere the phone is not.

## What we are working on

An Android build is the most requested thing and it is planned. Cloud-cover overlay is
the most requested feature, and it is the one we are most cautious about — cloud
forecasting is a different discipline with its own accuracy problems, and bolting a
weak version onto an honest aurora forecast would undermine the point of the app.
Better handling of substorm-scale variation is the change most likely to improve real
outcomes, because the three-hour Kp block genuinely hides good nights.

The city pages people keep asking for are coming to this site rather than the app: a
page per location with its geomagnetic latitude, the Kp it needs, its season window and
its realistic odds. We would rather publish forty of those properly than four hundred
thin ones.

## What we are not

We are not scientists and we do not present ourselves as an authority on
magnetospheric physics. The articles here are careful, sourced from public material,
and written by people who had to understand this well enough to implement it — which
is a real form of understanding, and also a limited one. Where the subject gets deep,
we point to the people who actually do the research rather than paraphrasing them
badly. If you are a space physicist and something here is wrong, we would genuinely
like the correction.

## Who this is for

Three kinds of reader, and the site is arranged around them. Someone who lives inside
the aurora zone and wants to stop missing displays that happen over their own house,
for whom the whole product is the alert. Someone planning a trip who needs to
understand which months and which latitudes make sense before spending money, for whom
the season and latitude tools matter far more than tonight's number. And someone
standing outside right now wondering whether to wait another twenty minutes, for whom
the only thing that counts is a current reading with an honest timestamp.

There is a fourth reader we are not trying to serve, and saying so saves everyone time:
if you want a curated feed of other people's aurora photographs, a social layer, or a
booking service, this is the wrong app and the wrong website. Several of the
alternatives on our [comparison pages](https://lumavik.org/compare/) do those things and do
them well.

## Privacy, briefly

The app has no accounts and no logins. Your saved locations and settings stay on your
device. If you turn on alerts, a coordinate rounded to half a degree and a push token
go to our relay so a notification can find you, and nothing else does. The tools on
this site run entirely in your browser and never send us what you type. The full
detail is in the [privacy policy](https://lumavik.org/privacy/), which is written to be read.

## Where to start

If you want the current conditions, the [live forecast](https://lumavik.org/tools/aurora-forecast/) is the fastest route. If you want to know whether your location is ever
realistic, run the [geomagnetic latitude calculator](https://lumavik.org/tools/geomagnetic-latitude-calculator/). If you want to understand the subject properly rather than
just check a number, the [Skywatch section](https://lumavik.org/skywatch/) is where we write it
all down, starting with [how accurate aurora forecasts actually are](https://lumavik.org/skywatch/aurora-forecast-accuracy/) — the piece that makes our own product look more
limited than any competitor's marketing, and the one we are most confident is true.

### The app

Free on iPhone. Forecast, oval map and one alert location at no cost, no account required.

[Get the app](https://lumavik.org/get/) Free · NOAA data · no account

---

HTML version: https://lumavik.org/about/
Structured data for this site: https://lumavik.org/api/v1/openapi.json · https://lumavik.org/llms.txt
Free to quote and reuse with a link back to the source URL above (CC BY 4.0).