Two engineers can build the same underwater optical wireless (UOWC) link — same blue-green laser, same photodetector, same data rate — and get wildly different results. One closes a comfortable 40 m link in clear ocean water; the other struggles to reach a few metres in a murky harbour. The hardware did not change. The water did. Understanding exactly how the water sets the range is the difference between a link budget that survives sea trials and one that fails on the first deployment.
This article explains the three impairments that govern every UOWC link — absorption, scattering and turbulence — how they combine, why water is classified the way it is, and what that means for range, wavelength choice and receiver design.
The three impairments — and why you must model all three together
An underwater laser beam is attacked in three distinct ways as it travels from transmitter to receiver:
- Absorption — water molecules, dissolved organic matter and (above all) chlorophyll convert photon energy to heat. This is usually the dominant loss mechanism and it is strongly wavelength-dependent.
- Scattering — suspended particles redirect photons out of the beam (and occasionally into it), spreading the beam and reducing the power that reaches the aperture.
- Turbulence — random fluctuations of the water's refractive index, driven by temperature and salinity microstructure, cause the received intensity to fluctuate (scintillation) and the beam to wander.
The common mistake is to budget for one and ignore the others. Absorption and scattering set the average loss; turbulence sets the fading around that average. A link can have plenty of average margin and still fail its availability spec because turbulent fades drop the instantaneous signal below threshold. A credible UOWC analysis multiplies all three effects into a single channel model.
Why UOWC is classified by chlorophyll concentration
Because absorption and scattering both scale with the amount of biological and particulate matter in the water, the community classifies natural water by chlorophyll concentration (mg/m³). Four canonical types span almost every real environment:
- Pure sea water — ~0.005 mg/m³. The clearest open ocean; absorption and scattering are at their lowest.
- Clean ocean water — ~0.31 mg/m³. Typical deep, open water.
- Coastal water — ~0.83 mg/m³. More biological activity and sediment.
- Harbour water — ~5.9 mg/m³. Heavily loaded with particulates and organics; the harshest common case.
The Beer–Lambert law ties this to loss directly: the received power falls as exp(−c(λ) L), where the attenuation coefficient c(λ) = a(λ) + b(λ) is the sum of the absorption and scattering coefficients, and L is the range. Because c rises steeply from pure sea to harbour water, the achievable range collapses in the same order — which is exactly why identical hardware behaves so differently from one deployment site to the next.
The blue-green window — and why 532 nm is a sweet spot
Water is far more transparent to blue-green light (roughly 450–550 nm) than to red or infrared, which is why every serious UOWC system operates there. But the optimum wavelength is a genuine trade-off, not simply "as blue as possible":
- Shorter (violet/blue) wavelengths minimise absorption in the very clearest water…
- …but turbulence-induced scintillation grows stronger as the wavelength shortens, and scattering behaviour also shifts with wavelength and water type.
In practice a green wavelength around 532 nm often gives the best all-round performance across clean, coastal and harbour water because it balances absorption against scattering and turbulence, rather than optimising for one alone. The "right" wavelength is the one that minimises the combined penalty for your specific water and range — which again is an argument for modelling all three effects at once.
Turbulence: from weak fades to saturation
UOWC turbulence is quantified by the scintillation index, which rises with the average temperature and salinity of the water and with a lower turbulent energy-dissipation rate (calmer, stratified water can actually be more turbulent, optically, than well-mixed water). The fading statistics are modelled with different distributions depending on strength:
- Log-normal for weak turbulence.
- Gamma–Gamma for moderate-to-strong turbulence.
One counter-intuitive but important effect is saturation: as turbulence becomes very strong, the scintillation index stops climbing and can even decrease slightly. Ignoring this leads to over-pessimistic BER estimates in the strongest regimes. A proper oceanic turbulence model — parameterised by real average temperature and salinity rather than an abstract index — captures this correctly.
What it costs you: the SNR budget by water type
The clearest way to see the impact is the SNR needed to hold a target bit-error rate. For a mid-range link (on the order of 15 m) aiming at a BER around 10−6:
- With turbulence alone, a modest SNR (roughly 20 dB) is enough.
- Add absorption and scattering, and the required SNR climbs sharply as the water degrades — from the high-20s of dB in pure sea, into the 40s of dB in clean ocean, and into the 60s of dB in coastal water.
- In heavy harbour water, the combined absorption and scattering can be so severe that the target BER is simply unreachable at that range for any practical SNR — the link must be shortened, or the requirement relaxed.
This is the quantitative reason UOWC is fundamentally a short-to-medium range technology — tens of metres in good water — and why a range that is trivial in clean ocean can be impossible in a harbour without moving the terminals closer together.
Aperture averaging: the cheapest turbulence mitigation you have
One design lever pays for itself more than almost any other: receiver aperture size. A larger collecting aperture averages the beam's intensity fluctuations across its face, so the scintillation seen by the detector — and therefore the BER — drops dramatically. Going from a millimetre-scale to a centimetre-scale aperture can improve the average BER by orders of magnitude under the same turbulence. Where the link geometry allows it, enlarging the aperture is often more cost-effective than raising transmit power. Beyond aperture averaging, spatial diversity (multiple receivers) and adaptive optics are the standard next steps for the strongest channels.
Designing the link before you get wet
The practical payoff of combining absorption, scattering and turbulence in one model is that you can answer the questions that actually decide a deployment — on a laptop, before committing to hardware or a sea trial: What range will a 532 nm link close in coastal water at my target BER? How much does doubling the receiver aperture buy me? Is my SNR budget realistic for harbour water, or do I need to shorten the link? Which modulation survives the fading at the available margin? Getting those answers up front is exactly where UOWC link-analysis software earns its place.
Frequently asked questions
Why is underwater optical communication limited to short ranges?
Because seawater absorbs and scatters light strongly. Even in clear ocean water the attenuation coefficient limits practical high-rate links to tens of metres; in coastal and harbour water — with far higher chlorophyll and particulate loads — the usable range drops to a few metres unless power, aperture or wavelength are optimised.
What wavelength is best for underwater optical wireless communication?
A blue-green wavelength in the ~450–550 nm window. The exact optimum depends on the water: the very clearest water favours shorter (bluer) wavelengths, but scattering and stronger turbulence scintillation at short wavelengths mean a green wavelength around 532 nm is frequently the best all-round choice across clean, coastal and harbour water.
How does water type affect a UOWC link?
Water is classified by chlorophyll concentration — from ~0.005 mg/m³ in pure sea to ~5.9 mg/m³ in harbour water. Higher chlorophyll means more absorption and scattering, a larger attenuation coefficient, and therefore shorter range and a higher required SNR for the same bit-error rate.
Does a bigger receiver really help underwater?
Yes. A larger receiver aperture averages out turbulence-induced intensity fluctuations, reducing scintillation and improving the average BER by orders of magnitude in many cases. Aperture averaging is one of the most cost-effective turbulence mitigations available for UOWC.
SCOPT builds physics-based tools for exactly this analysis — modelling absorption, scattering and turbulence together across pure-sea, clean-ocean, coastal and harbour water, with BER, SNR and scintillation predictions versus range, wavelength, water parameters and receiver aperture. Explore our software tools, our companion guide on underwater optical turbulence, or request a consultancy.