A Fibre Laser Built For Wind

Multiply the two numbers in the headline and you get three watts. That is the published target for the laser module Laser Zentrum Hannover announced on 30 September: pulses of around 100 microjoules, a repetition rate of at least 30 kilohertz, an average output power of at least three watts. Three watts is not an impressive figure on its own — a kilowatt-class industrial fibre laser does several hundred times more. The difficulty is entirely in how those three watts are divided up, and in the fact that the box holding them has to survive on a mast, on a roof, or beside a runway for years without a service visit.

The module is being built inside VeLiS — Versatile LiDAR System — a project LZH is running with Thies Clima, the Göttingen meteorological instrument maker. LZH is doing the laser. Thies is doing the electronics, the evaluation algorithms and the lidar as a whole. The work is funded by Investitions- und Förderbank Niedersachsen (NBank) out of European Regional Development Fund money; Thies adds that it runs under Lower Saxony's innovation support programme for corporate research and development (IFP), with funding from the State of Lower Saxony as well. The partners have been talking about it publicly since Thies announced the project on 17 August.

A cup anemometer tells you the wind at the anemometer. That is its entire limitation, and it is a serious one: a wind farm developer wants the wind at hub height across a site, an airport wants the wind along an approach path, and a grid operator wants to know about the gust that has not arrived yet. None of those places has an instrument in it.

A Doppler wind lidar fires a pulse into clear air and listens to what comes back off the aerosol — dust, salt, pollen, combustion particles, the haze that is always there. Those particles move with the air, so the light they scatter returns frequency-shifted by the component of the wind along the beam. Measure the shift and you have one velocity component at whatever range the return came from. Point the beam in three or more directions and you can reconstruct a wind vector at every range gate along each line of sight. The instrument never touches the air it measures.

Aerosol backscatter is desperately weak. The fraction of transmitted light that finds its way back to the telescope from a kilometre of clear air is small enough that the returning signal is counted in photons, not milliwatts, and it falls off with the square of range on top of that. The number of photons in a single return is set by the energy in the pulse that produced it. Double the pulse energy and you double the return from every range gate — which, in a measurement limited by shot noise, buys a useful extension of the range at which a single shot still means something.

Repetition rate buys something different. The signal from one pulse at long range is usually below the noise, so a wind lidar averages: thousands of shots are accumulated before a velocity is extracted. Averaging n shots improves the signal-to-noise ratio as the square root of n, and n is the repetition rate multiplied by the dwell time. A source that fires twenty-five times more often therefore reaches a given confidence twenty-five times sooner, or delivers five times the signal-to-noise ratio in the same dwell. Rate is how quickly the instrument can revisit a direction; energy is how far it can see at all.The two pull against each other, because their product is average power, and average power is what a compact, passively cooled, mast-mounted enclosure has to get rid of. LZH's three watts is stated as a floor, not a ceiling — "at least three watts" — but in a sealed field box without active cooling the thermal headroom above that figure is not large, and the designer's real freedom is in where on the energy-versus-rate curve those watts are spent. VeLiS has picked an unusually energetic point, and LZH says so plainly: laser modules currently on the market, it notes, offer significantly lower pulse energies at comparable repetition rates.

A 100-microjoule pulse at 30 kilohertz is hard to make in fibre for a reason that has nothing to do with the pump and everything to do with the last metre of glass. Pulse energy divided by pulse duration is peak power, and peak power in a fibre core of a few hundred square micrometres is an intensity high enough to turn the glass itself into an active participant. Stimulated Brillouin scattering reflects the pulse back down the amplifier. Self-phase modulation smears its spectrum — fatal in a coherent instrument, where the whole measurement is a frequency comparison and the transmitted linewidth sets the velocity resolution. Four-wave mixing and stimulated Raman scattering move power to wavelengths the receiver is not looking at.

The usual escapes are to lengthen the pulse, enlarge the mode area, or shorten the amplifier. Each costs something: a longer pulse coarsens the range gate, since the gate length is roughly half the pulse length in air, the light having made a round trip; a larger mode area makes the fibre stiffer, more bend-sensitive and harder to keep single-mode; a shorter amplifier needs more heavily doped glass and more pump brightness. The engineering of VeLiS's module will be a negotiation among those three, and neither partner has yet said where it will land — no pulse duration, no wavelength, no linewidth has been published, and without those the 100-microjoule figure cannot be fully read. Coherent wind lidar is conventionally built around 1.5 micrometres, where eye-safety limits are generous and telecom components are cheap, and nothing in the announcement suggests VeLiS departs from that.

Read that sentence as a hardware specification rather than a press-release flourish and it is the harder half of the project. LZH's stated requirements for the module are a compact, energy-efficient design, the least possible weight, robustness against vibration and environmental influences, and reliable operation with minimal maintenance. A wind lidar on a met mast lives through large seasonal and diurnal temperature swings, condensation, salt if it is offshore, and vibration at the tower's own resonances. Fibre is good at this; free-space alignment is not, which is most of why coherent wind lidar went all-fibre in the first place. But a 100-microjoule amplifier has splices, a pump combiner, a large-mode-area gain fibre and an end cap that all have to hold their alignment and their damage margin through every one of those cycles, and the end facet is the usual place where field units die.

Thies Clima's side — the electronics, the evaluation algorithms and the system integration — carries the other half of the cost argument. A wind lidar's bill of materials is not dominated by the laser; it is dominated by the laser plus the receiver plus the scanner plus the enclosure plus the compute, and the reason commercial units cost what they do is that all five are specialised. A modular set of technology building blocks that can be reconfigured for a flight corridor, a long-term site assessment, or predictive gust detection beside energy infrastructure is an attempt to spread that fixed cost over three markets instead of one.What has changed

Not very much, yet. VeLiS is a development project with a target specification, not a product with a datasheet, and the interesting numbers — pulse duration, linewidth, wall-plug efficiency, the range it actually achieves in a real atmosphere — are the ones still missing. Nor has either partner put a named engineer behind the figures; both announcements are institutional, with no quoted project lead. What the announcement does mark is where a serious laser institute now thinks the gap is. Not in average power, which fibre has had for twenty years. Not in beam quality. In the unglamorous combination of pulse energy, repetition rate and survivability that a field instrument needs, delivered at a price a weather-instrument company can put in a catalogue. If VeLiS lands, the component that gets cheaper is not the laser — it is the wind measurement, and the people who notice will be the ones currently paying for a met mast.