LAB-005 · methodology
Why We Listen at 1420 MHz
Dee Okafor ·
The hydrogen line sits at 1420.405750 MHz, give or take a few hertz depending on the source’s radial velocity relative to us. That frequency is not a choice anyone made. It is the natural emission frequency of neutral hydrogen when a proton captures an electron and the electron flips its spin from parallel to antiparallel. The energy gap in that transition is small, and the event is rare for any given atom, but there is more neutral hydrogen in the galaxy than anything else, so the signal is everywhere: faint background noise from the galactic plane when you point at it, a specific spectral peak when you integrate long enough.
What makes 1420 MHz useful for SETI work rather than just general radio astronomy is two things together. First, it is a protected band. The ITU designates 1400–1427 MHz passive only: no licensed transmitters permitted in that range, which keeps it comparatively free from strong in-band interference. Second, any civilization doing radio astronomy would almost certainly know this frequency. It is one of the most obvious spectral markers in the observable universe. Choosing a frequency a receiving civilization might also be monitoring is a defensible prior, though not a guarantee of anything.
Bernard Oliver coined the water hole concept as part of Project Cyclops in the early 1970s: the quiet stretch between the hydrogen line at 1420 MHz and the hydroxyl radical (OH) line at 1662 MHz. His framing was that these two lines bracket the spectrum’s most chemically natural window. H and OH together, the components of water. Whether that chemical logic matters to any actual sender is unknown, but it is a reasonable place to listen when you have no better information. Frank Drake’s account of how the SETI community arrived at that consensus is the most readable version of this reasoning: he co-wrote Is Anyone Out There? with Dava Sobel, which covers Project Ozma and the methodology behind the water-hole choice. Drake presents the water hole as a working assumption, not a derivation, which is the right level of confidence for a field with a forty-year null-result record.
We adopted that consensus because it is the only principled prior we have. The interference environment at L-band is as good as anything available for a fixed ground installation, and the LNA stage of our receive chain is built around it. The SAWbird+ H1 is a surface-acoustic-wave-filtered low-noise amplifier tuned to the hydrogen line. SAW filtering at the front end rejects out-of-band interference before it reaches the mixer stage. For weak-signal work at the expected flux levels, that filter is not optional.
If you are building toward 1420 MHz reception from scratch: an RTL-SDR V3 gives you a working entry point. Its dynamic range is the limiting factor at the entry tier, but you will learn what the band actually sounds like before you need to spend more. Pair it with the SAWbird+ H1 LNA directly behind your antenna, and budget more for coax quality than feels intuitive — feedline loss at 1420 MHz over a long run between the dish and the receiver will cost you more than a better antenna would have gained. The full receive chain is in LAB-006; this note is specifically about the frequency rationale, not the hardware stack. Once you have a spectrum, the Doppler calculator converts what you measured against 1420.405751 MHz into velocities, corrected to the local standard of rest. The rest of our calculators are on the tools page.
One practical note on the protected-band claim: passive-only designation removes licensed transmitters, not all interference. Consumer electronics and switching power supplies generate incidental radiation across L-band. The ITU protection is most useful for eliminating strong in-band interferers; the noise floor you can reach beyond that depends on shielding, cable quality, and how much RFI you are generating in the equipment room yourself.
Walt’s running list of reasons the signal is not terrestrial noise goes back to the early sessions. The first item is still the frequency: the persistent signal is in-band, centered correctly relative to the hydrogen line, and its structure does not match the incidental interference catalogue we have built up over the years. That is a necessary condition, not sufficient, but it is the right place to start.
The links to the SAWbird+ H1 and the RTL-SDR V3 above are affiliate links. If you were going to buy either one, buying through those links offsets a portion of what this project costs to run, and we would rather disclose that plainly than not.