Tag: NOAA Teacher at Sea

NOAA Teacher At Sea

Staci DeSchryver

Aboard NOAA Ship Oscar Elton Sette

July 6 – August 2, 2017

Mission:  HICEAS Cetacean Study

Geographic Area:  Near the Maro Reef, Northwest Hawaiian Islands

Date:  July 24, 2017

Weather Data from the Bridge:

Location: 23 deg, 39.5 min N, 169 deg, 53.5 min W

Wind:  85 degrees at 12 kts

Pressure:  1017.0

Waves: 2-3 feet at 95 degrees

Swell: 3-4 feet

Temperature 27.5

Wet bulb temp: 26.2

Science Log

Most of us know the first rule of Fight Club – Don’t talk about Fight Club.  In previous blogs, we’ve established that if acoustics hears a vocalization from the lab, they do not inform the observers on the flying bridge – at least not until all members of the vocalizations are “past the beam”, or greater than 90 degrees from the front of the ship.  Once the vocalizations are past the beam, acoustics can elect to inform the observers based on the species and the specific protocols set for that particular species.  The purpose of this secrecy is to control for bias.  Imagine if you were a marine mammal observer, headed up for your last two hour shift on your ten hour day.  If you stopped by the acoustics lab to say hello and found the acoustician’s computer screens completely covered with localizations from a cetacean, you might change the way you observe for that animal, especially if you had a general idea of what angle or direction to look in. One experimental goal of the study is to eliminate as much bias as possible, and tamping the chatter between acousticians and the visual team helps to reduce some of this bias.  But what about the observers?  Could they bias one another in any way?  The answer to that question is yes, and marine mammal observers follow their own subset of Fight Club rules, as well.

Let’s say for example, a sighting of Melon-Headed Whales is occurring.  On the flying bridge, available observers come up to assist in an abundance estimate for that particular group (more on how these estimates are made later).  They also help with photographing and biopsy operations, when necessary.  Melon-Headed Whales are known to travel in fairly large groups, sometimes separated into sub groups of whales. After spending some time following the group of whales, the senior observer or chief scientist will ensure that everyone has had a good enough opportunity to get a best estimation of the number of Melon Headed Whales present.  At this point, it’s time for the observers to write their estimates.  Each observer has their own “green book,” a small journal that documents estimation numbers after each observation occurs.  Each observer will make an estimation for their lowest, best, and highest numbers.  The lowest estimate represents the number of cetaceans the observer knows for certain were present in the group – for example they might say, “There couldn’t possibly be fewer than 30”.  The highest estimate represents the number that says “there couldn’t possibly be any more than this value.”  The best estimate is the number that the observer feels totally confident with.  Sometimes these values can be the same.  The point is for each observer to take what he or she saw with their own eyes, factor in what they know about the behavior of the species, and make a solid personal hypothesis as to the quantitative value of that particular group.  In a sighting of something like our fictitious Melon Headed Whales, those numbers could be in the hundreds.

Once the documentation is complete in the green books, the observers direct the ship to return back to the trackline, and begin observing again.  They never discuss how many animals they saw.  This is such an important part of what marine mammal observers do as professionals.  At first glance, one would assume that it would be beneficial for all observers to meet following an observation to come to a consensus on the numbers sighted.  But there are a lot of ways that discussion on numbers can turn sideways and skew overall data for the study.  Let’s take an obvious example to highlight the point.

Imagine if you were a new scientist in the field, coming to observe with far more senior observers.  Let’s assume you’ve just spotted a small group of Pygmy Killer Whales and although you are new on the job, you know for an absolute fact that you counted six dorsal fins – repeatedly – through the course of the sighting.  If the sighting ends, and the more senior observers all agree that they saw five, the likelihood that you are going to “cave” and agree that there were only five could be higher.

If you never talk about your numbers, you never have to justify them to anyone else.  The question often comes up, “What if an observer consistently over or underestimates the number of cetaceans?”  It’s much better for the scientists to consistently over or underestimate their counts than to spend time trying to fine tune them against the rule of another’s estimate.  If counts skew high or low for a scientist each leg of the trip as the co-workers change, that can create a problem for those trying to analyze the abundances after the study is complete.  Further, not discussing numbers with anyone at all ever gives you a very reliable estimation bias over time.  In other words, if you consistently over estimate, the people who complete the data analysis will know that about you as an observer and can utilize correction factors to help better dial in cetacean counts.  It is because of this potential for estimation bias that all marine mammal observers must never talk numbers, even in casual conversation.  You’ll never hear a marine mammal observer over dinner saying, “I thought there were 20 of those spinner dolphins, how many did you think were there?”

Where do these data go after the study is over?  Data from each sighting gets aggregated by the chief scientist or other designee and the group size for each sighting is determined.  Then, via many maths, summations, geometries, and calculuses, population abundance estimates are determined.  This is a dialed-in process – taking the number of sightings, the average sighting group size, the length of the transect lines, the “effective strip width” (or general probability of finding a particular cetacean within a given distance – think smaller whales may not be as easy to see from three miles away, and therefore the correction factor must be taken into account), and finally the probability of detection – and combining those values to create a best estimate for population density within the Hawaiian EEZ.

The probability of detection is an interesting factor in that it used to always be considered as a value of 1 – meaning that if a cetacean shows his friendly (or ferocious) mug anywhere on the trackline (the predetermined path the ship is taking in the search) the value assumes that a mammal observer has a 100% chance of spotting it.  This is why there is a center observer in the rotation – he or she is responsible for “guarding the trackline,” providing the overlap between the port and starboard observers in their zero to ninety degree scans of the ocean.  Over time, this value has created statistical issues for abundance estimates because there are many situations when a 100% detection rate is just not a realistic assumption.  Between the HICEAS 2002 study and the HICEAS 2010 study, these detection factors were corrected for, leading to numbers that were reliable for the individual study itself, but not reliable to determine if populations were increasing or decreasing.

Other factors can play a role in skewing abundance estimates, as well.  For example, beaked whales often travel in smaller-sized groups and only remain at the surface for a few minutes before diving very deeply below the surface.  Sightings are rare because of their behavior, but it doesn’t necessarily mean that they are declining in population.  In HICEAS 2002, there was an unusual sighting of a large group of these whales.  When the statistical methods were applied for this group as a whole, the abundance numbers were very high.   In 2010, the sighting frequency was more “normal” than finding the anomalous group, and the values for the numbers of these whales dropped precipitously.  There wasn’t necessarily a decline in population, it just appeared that way because of the anomalous sighting from 2002. Marine mammal observer Adam Ü assists on a sighting by taking identification photos.

Statistical analysis methods have also changed over the years once scientists took a harder look at some of the variables that the marine mammal observers must contend with in their day to day operations.  At the start of every rotation, mammal observers make general observations about the sea conditions – noting changes in visibility, presence of rain or haze, wind speed, and Beaufort Sea State.  Observers will go “off effort” if the Beaufort Sea State reaches a 7.  To give you an idea of how the sea state changes for increasing numbers, a sea state of Zero is glass-calm.  A sea state of 12, which is the highest level on the Beaufort scale, is something I’m glad I won’t see while I’m out here.  Come to think of it, we have gone “off effort” when reaching a sea state of 7, and I didn’t care for that much, either.    

Most of our days are spent in at least a Beaufort 3, but usually a 4 or 5.  Anything above a 3 means white caps are starting to form on the ocean, making it difficult to notice any animals splashing about at the surface, especially at great distances – mainly because everything looks like it’s splashing.  Many observers look for splashing or whale blows as changes against the surrounding ocean, and the presence of waves and sea spray makes that job a whole heck of a lot more difficult.  Beaufort Sea States are turning out to be a much bigger player in the abundance estimate game, changing the statistical probabilities of finding particular cetaceans significantly.  

One species of beaked whale has a probability of sighting that drops off exponentially with increasing sea state.  As sea state goes up, the chances of seeing any cetacean at all decreases.  Other factors like sun glare play a role in decreased sightings, as well.  When a beaked whale “logs” at the surface in glass calm waters, chances are higher that it will be spotted by an observer. When the ocean comes up, the wind is screaming, and the waves are rolling, it’s not impossible to see a whale, but it sure does get tough.

The good news is that for most species, these abundance estimates account for these variables.  For the more stealthy whales, those estimates have some variation, but overall, this data collection yields estimate numbers that are reliable for population estimates.

Personal Log

It is darn near impossible to explain just how hard it is to spot mammals out in the open ocean.  But, being the wordy person I am, I will try anyway.

I had some abhorrently incorrect assumptions about the ease at which cetaceans are spotted.  These assumptions were immediately corrected the first time I put my forehead on the big eyes.  Even after reading the reports of the number of sightings in the Hawaiian EEZ and my knowledge of productivity levels in the tropical oceans,  I had delusions of grandeur that there would be whales jumping high out of the water at every turn of the ship, and I’d have to be a blind fool not to see and photograph them in all of their whale-y glory.

I was so wrong.

Imagine trying to find this:

In this:

Here’s the long and short of it – there were times when we were in pretty decent conditions, and marine mammal observers were “on” a sighting, and I trained the big eyes in exactly the direction and my eyes at the exact distance and I still couldn’t see them.  There were times when the mammals pretty much had to be launching themselves out of the water and onto the ship before I was like, “Oh, hey!  A whale!”  I can think of at least four sightings where this happened – whales were out there, everyone else could see them…and I couldn’t find them if they were pulled out of the water and handed to me in a paper bag.  Which is extra disappointing because a) a whale doesn’t fit in a paper bag, and 2) if it did, it would likely soak the bag so that it fell out of the bottom and now I’d have a whale that I couldn’t see anyway who now has a headache and is ornery because someone shoved him in a paper bag that he promptly fell face first out of.  And as I’ve learned over the time I’ve been on the ship and through many forays into the wilderness – don’t anger things with teeth.

I have had the good fortune of watching our six marine mammal observers as they do their work and I am continually floored at the ability and deftness in which they do their jobs.  I have done a few independent observation rotations – I try to get in at least three each day – and I have only once been able to complete a rotation in the same way the observers do.  Looking for forty minutes through the port side big eyes, sitting and guarding the trackline for 40 minutes, and looking for forty minutes through the starboard side big eyes is exhausting.   Weather conditions are constantly changing and sometimes unfavorable.  The sun could be shining directly in the path of observation, which turns the whole ocean into the carnage that could only be rivaled by an explosion at a glitter factory.  While the canopies protect the observers from a large majority of incoming sunlight, there’s usually a few hours in the day where the sun is below the canopy, which makes it blast-furnace hot.  Today the winds are blowing juuuuust below the borderline of going off effort due to sea state conditions.  Sometimes the wind doesn’t blow at all, or worse –  it blows at the exact speed the ship is traveling in – yielding a net vector of zero for wind speed and direction.  Out on the open ocean, Beaufort Sea States rarely fall below a 3, so observers are looking through piles of foam and jets of sea spray coming off the waves, searching for something to move a little differently.  Trying to look through the big eyes and keep the reticle lines (the distance measures on the big eyes) on the horizon during the observation while the ship moves up and down repeatedly over a five foot swell?  I can say from direct experience that it’s really, really hard.

The animals don’t always play nice, either.  It would be one thing if every animal moved broadside to the view of the observers, giving a nice wide view of dorsal fin and an arched back peeking out of the water.  A lot of cetaceans see ships and “run away.”  So, now as an observer, you have to be able to spot the skinny side of the dorsal fin attached to a dolphin butt.  From three miles away.   Some whales, like sperm whales, stay at the surface for about ten minutes and then dive deep into the ocean for close to an hour.  We’re lucky in that if we aren’t on the trackline and spot their telltale blows when they are at the surface, the acoustics team knows when they are below the surface and we can wait until they do surface, so that’s a benefit for everyone on the hunt for sperm whales.

But overall? These things are not easy to find.   We aren’t out here on a whale watching tour, where a ship takes us directly out to where we know all the whales are and we have endless selfie opportunities.  The scientific team couldn’t bias the study by only placing ourselves in a position to see cetaceans.  In fact, the tracklines were designed years ago to eliminate that sort of bias in sampling.  Because we cover the whole Hawaiian EEZ, and not just where we know we are going to see whales (looking at you, Kona) there could be times where we don’t see a single cetacean for the whole day.  As an observer, that can be emotionally taxing.

And yet, the marine mammal observers persevere and flourish in this environment.  Last week, an observer found a set of marine mammals under the surface of the water.  In fact, many observers can see mammals under the water, and it’s not as though these mammals are right on the bow of the ship – they are far far away.  Most sightings happen closer to the horizon than they do to the ship, at least initially.  The only reason why I even have pictures of cetaceans is because we turn the ship to cross their paths, and they actually agree to “play” with us for a bit.   

Over the last three weeks, I’ve tried to hone my non-skill of mammal observation in to something that might resemble actual functional marine mammal observation.  I have been thwarted thus far.  But I have gotten to a certain point in my non-skill – where at first, I was just in glorious cod-faced stupor of witnessing cetaceans, and trying to get as many photos as possible – now, a sighting for me yields a brief moment of awe followed by an attempt to find what the observers saw in order to find the animal.  In other words, I “ooh and ah” for a few moments at first, but once I can find them, I start asking myself, “Ok, what do the splashes look like?”  “How do the fins look as they come out of the water?”  “What does the light look like in front or behind the animal, and would I be able to see that patterning while I’m doing an observation?”  So far, I’ve been unsuccessful, but I certainly won’t stop trying.  I have to remember that the marine mammal observers who are getting these sightings have been doing this for years and I have been doing this for hours comparatively.  Besides, every sighting is still very exciting for me as an outsider to this highly specialized work, and the star-struck still hasn’t worn off.  I imagine it won’t for quite some time.  

Ship Fun!

Being at sea for 28 days has its advantages when it comes to building strong connections between scientists, crew, and the officers.  Everyone pitches in and helps to make life on this tiny city a lot more enjoyable.  After all, when you spend 24 hours a day on a ship, it can’t all be work.  Take a look at the photos below to see: