The Best Australian Science Writing 2015 (15 page)

Authorities are doing additional mapping of the area with three modern sonar survey ships, including one contracted from Fugro NV. But their equipment operates from the surface, and the images are unlikely to be sharp enough to spot plane debris, instead revealing the lay of the land for the next group of searchers being brought in under the fresh tender.

To find the plane, those searchers will have to use undersea sonar equipment, including side-scan sonars, that can be sent to within 50 metres of the seabed.

Coordinating the search tender from a drab office block in Canberra is the ATSB, and its chief commissioner, Martin Dolan, a long-time public servant who previously ran Australia's workers' compensation program. Helping him sift through search bids is a panel of five senior bureaucrats, including a former BP PLC ocean-fleet manager and an ATSB marine engineer.

The decision will ultimately be Mr Dolan's, and the bidding rules allow him to consider more than merely the lowest bidder. There will be some tough choices, which could be crucial in determining whether contractors fulfil their mission of finding the plane. Among the hardest: whether to bet heavily on companies like Williamson & Associates, which specialises in using relatively simple towed sonars, or outfits like Woods Hole that have their own higher-tech autonomous underwater vehicles, or AUVs. A wrong choice could reduce the search's odds of success.

Towed sonars are attractive because they can cover large areas relatively quickly – an important advantage given that ATSB officials are giving bidders just 300 days to explore the 32 000 square kilometre search area due to intense public interest.

But they are also sometimes hard to manipulate. Searching for the AHS
Centaur
, a World War II-era hospital ship, in 2009, scientists had to tow a sonar device with more than 6 kilometres of cable through a ravine 90 metres wide. They likened it to the movie
Star Wars
, when fighters flew through a narrow trench at high speeds to unload a final shot to destroy the fictional Death Star. One device snapped off its cable and was lost.

The AUVs, by contrast, can be programmed to navigate finicky terrain. But AUVs cover less ground because their batteries can't power energy-intensive sonars, and they cannot be operated in rough seas as they have to be hauled back on deck each day, placing a ship's crew at risk.

They also can go haywire. Woods Hole's flagship deep-sea AUV, the
Nereus
, imploded under high pressure off the coast of New Zealand in May. Another multimillion-dollar AUV, the REMUS 6000, failed to resurface on a recent research trip. ‘The decision, fortunately or unfortunately, in the end rests with me,' said Mr Dolan.

Most of the firms that spoke with
The Wall Street Journal
said they were bidding as part of consortia, as very few have the
equipment, ships and expertise to carry out the massive search alone. A number have already descended upon Canberra to make their cases, including Mr Wright of Williamson & Associates, who along with Mr Mearns of Blue Water Recoveries worked earlier with Australian officials to locate a torpedoed World War II Australian warship. Ms Keller of Metron said she would relocate to Canberra to operate a data centre crunching sonar images from search ships and updating authorities, if her firm wins the bid.

Once the search is finished, if the whole seabed ends up being mapped, the area will be the most comprehensively interrogated strip of ocean in the world, Mr Dolan of the ATSB said. Scientists will know every boulder and divot that dots its surface. Whether that includes Flight 370 remains to be seen.

Postscript:
Following the completion of tender arrangements, Fugro NV were awarded the contract to carry out the search for MH370 alongside a team hired by Malaysia. In April 2015 Australia, Malaysia and China committed to doubling the search zone to 120 000 square kilometres, an area roughly the size of England. At the time of publication, more than 48 000 square kilometres of this part of the remote Indian Ocean sea floor had been scanned with high-definition sonar devices, turning up new volcano formations, a shipwreck, but no sign of the missing plane.

All dressed up for Mars and nowhere to go

Robots on a roll

Field guide to the future

Ian Lunt

I could ask, ‘what was your first field guide?' but my first field guides belonged to my parents, not me. So instead I'll ask, ‘what was the first field guide you remember using?'

I remember two:
Trees of Victoria
by Leon Costermans – a permanent resident of the car glove box – and
Birds of the Ranges
by the Gould League. I am indebted to the authors and illustrators of both. Without them, I might have led a different life.

Our Costermans bore the hallmark of a truly great field guide; after years of abuse, we stripped it of every skerrick of resale value. One summer, someone put a block of copha in the car glove box to protect it from the sun. When discovered weeks later, everything floated in a pool of coconut fat. It was awesome. Costermans was indestructible. Like the trees inside it, we created the world's first rip-proof, waterproof, scented and highly combustible field guide to eucalypts.

* * * * *

What makes a field guide truly great? There are four rules. Every great guide must be: (1) distinctive, (2) attractive, (3) well organised, and, most important of all, (4) impeccably accurate. As in
life, one small mistake about eye colour (‘Your eyes are a beautiful brown'. ‘Aah, they're green actually'), and into the reject bin we fall.

Many of the great guides are also (5) exhaustively comprehensive, but that rule isn't rigid. Comprehensiveness is equally an asset and a liability. A good field guide can't be huge; it has to fit in the backpack.

* * * * *

Our generation witnessed and abetted the great migration. The migration of information from paper, plastic, canvas and vinyl, to invisible strings of zeroes and ones. The migration to digital. Never despair, books won't disappear; familiarity breeds contempt for usurpers. It's hard to find content, let alone contentment, in a replacement; especially one with no pages.

Yet resist as we will, the next generation of great field guides won't be on paper, but on a device. The
Compleat Naturalist
of the future will treasure a phone more than a library. To my offspring, a bird in the handset is worth three in a book.

* * * * *

What will we demand from the digital field guides of the future? That they be distinctive, attractive, well organised, accurate? These maxims won't change, but one rule will vanish and a new rule will emerge. Comprehensiveness will no longer be a liability. Size won't matter, as the
longest
field guide need be no
bigger
than the shortest. Every entry in a digital field guide can expand and contract like a concertina, from mug-shot to museum and back again.

As encyclopedias shrink in the palm of our hand, the field guides of the future lie prone to a new imperative:
be popular.
In
the olde days, greatness led to popularity. In the digital world, popularity leads to greatness; not through imitation, but through content creation. As each user adds and shares new information, the most popular guides will blossom and grow, while the wallflower guides wither and die.

* * * * *

Many field guides have migrated from paper to digital. Some synced smoothly from book to app. Others grew in stature as they shrunk in size. Augmented by libraries of recorded calls, the better guides to frogs and birds can now be viewed, and heard, on a mobile phone. Can't tell a
Crinia signifera
from
Crinia parinsignifera
? The pictures won't help, but the calls certainly will.

Calls aside, this first generation of digital field guide is characterised by the ‘silent cinema syndrome'. Remember the first silent movies? How quaint and silly they appear? The pioneers of the silent movie transported an old aesthetic to a new technology. They pushed the stage show onto the silver screen. But they created nothing new, beyond a distribution network. They had no concept for the
cinematic.
The art of cinema evolved later, as unruly directors liberated film from the stagecraft of the theatre. The evolution of cinema was enabled by technology and directed by new worlds of imagination.

A century beyond the birth of cinema, we squeezed great books onto the small screen. We added a soundtrack. But we are yet to generate a new genre, a new art form, worthy of the title, the Truly Great Digital Field Guide.

* * * * *

A young phone has a great eye, perfect ears, superb short-term memory, an encyclopedic brain, and looks fantastic.

Too anthropomorphic? A little bit creepy? Let's try again. A modern phone has a sensitive microphone, quality camera, accurate voice recorder and connects to the cloud. Thanks to a GPS, it knows exactly where it is most of the time, which is more than I can claim. The fab guides of the future will feast on every one of these features. Why? Check out my apps.

Hear a song you don't know? Grab your phone, point to the sound and – Shazam – the name of the song, artist and album appears in a flash, with links to biography, lyrics and videos. It's magical. My phone records a sound bite, sends it to a satellite, compares it to a database, and sends me back the artist's face.

If this is possible, then why do I have to scroll manually through frog call after frog call to work out which critter is croaking in the dam? Surely a Truly Great digital field guide should tell me who made the call?

* * * * *

What about that camera? Leafsnap is a groundbreaking app for identifying plants in north-east America. Point your phone at a plant, take a photo of a leaf, beam it to the cloud, and it shoots back a list of potential species, with additional photos of fruits, flowers and distinguishing features. There's no need to measure the leaf, or to look up a glossary to determine if it's palmate or pinnate, just point, snap and leave the work to the app.

Every meat-eater knows that Leafsnap has limited potential; most leaves look the same to a carnivore. But picture a smorgasbord of photo ID apps, like FruitSnap, BudSnap, BarkSnap and EucSnap. Australia has over 700 species of
Eucalyptus
, all with distinctive fruits (‘gum-nuts'), buds and bark. Two or three photos should be all it takes for EucSnap to identify most trees. Photograph a fruit, send it to the cloud, and receive a shortlist of possible species, filtered by locality. Snap some buds, send them
back up, and the shortlist shrinks again. Can't tell your Snugglepot from Cuddlepie? Just shoot a gum-nut to the sky.

* * * * *

Shazam nails song identification. It gets ‘Hurt' by Nine Inch Nails every time I play it. It knows its covers, and gets ‘Hurt' by Johnny Cash every time too, just like me. But bird songs aren't pop songs. Both versions of ‘Hurt' always sound the same. The crimson rosella outside my window doesn't.

A bird is an artist with an improvised repertoire. Blackbirds sing more songs than the Beatles ever sang. To contain a bird, a library must call upon a discography, not a whistle. To make it harder, blackbirds ain't blackbirds. Birds have accents.

* * * * *

In 2011, a science reporter at the University of Wisconsin wrote an enthusiastic press release about WeBIRD; a phone app to let users, ‘record a nearby bird's call, submit that song wirelessly to a server and retrieve a positive ID of the species'. Shazam.

Apparently, the prototype identified the resident birds of Wisconsin really well. Users could upload a call and download an ID, as the press release promised. But beyond Wisconsin, WeBIRD went wonky, as many birds sing different tunes in different towns. WeBIRD couldn't cope with accents. Its vocal files were too localised. It needed a bigger jukebox.

The solution? Collect more songs and data files and try again. Surely, centuries of natural history should provide reams of data to train a smart app? Yes and no. We've got piles of information. It's just the wrong type. Or rather it's
all type.
It's all text.

* * * * *

Words are cheap. Since birth, natural history has been subject to this economic dictum of the printing press. Images and sounds are expensive to reproduce, so generations of biologists created ways to translate images and sounds into words.

Botanists penned glossaries to describe undrawn plants: ‘
leaves conduplicate, seeds tuberculate, indumentum villose, calyx accrescent
'. The efficiency of communication is exemplary.
A Handbook to the Plants of Victoria Volume II. Dicotyledons
, published in 1972, distinguishes 2290 species in 832 pages without a single picture. Unburdened by art, both volumes – plus lunch, water, camera and coat – fit snugly in my day-pack.

With a budget for paint – one illustration per species – but none for sound, cheerful ornithologists turned to onomatopoeia: ‘
Pee-pee-pee-peeooo, Wee-willy-weet-weet, It-wooa-weet-sip, Zzzt zzzt zzzt. Cher-cher-cherry-cherry, Wah-i-wah-i-wah-oo, Twitchy tweedle, Kupa-ko-ko, Lik-lik-lik
'. Less cheerful colleagues followed suit: ‘
Chop-chop, Four o'clock, Wide-a-wake, Walk to work. Want a whip? It's for teacher. Tweet-your-juice, Sweet pretty creature
'. (All real calls, I assure you.)