Destination Moku o LoʻeLatest update June 13, 2019 Started on March 22, 2019
A summer course at Coconut Island (Moku o Loʻe) in Kaneohe Bay, Hawaii, will focus on innovative approaches to marine science. We will use drones, photogrammetry, laser scanners, and remote sensing to explore the marine world.
This course was a game-changer. I have always known I would become a marine scientist who researched marine mammals (cetaceans, mostly), and that I was interested in our ever increasing knowledge and use of technology within that science. Beyond that? I wasn't really sure where to go, how to get there, or what was available. Applying Innovative Technologies in Marine Science answered those questions and raised maybe 100+ more.
I have been working with Whalength, the software designed specially for gathering morphometric data of whales, on multiple species around the world for over 6 months. Yet, that is only a piece of the extensive work being done by the scientists running the projects. This course was able to let me see the full circle of the research, beginning to end. I got up-close encounters with the Unoccupied Aerial System (UAS) used to take the videos, I pulled the still images and altitudes from the videos previously taken, I plugged those into Whalength (my true, tried, and trusty friend), analyzed the images, and used the data gathered to calibrate physical to UAS measurements and determine change in volume for multiple animals of differing species, locations, and season. Being able to see the time and effort that goes into working with just a small sample size gave me great appreciation for the teams that put in the work, but it also made me excited to see where I can go with this type of research. I also was given the opportunity to work with people from around the world and make connections that I know will last me many years to come, and potential collaborations with future research. I can't thank the instructors of this course enough for giving me this chance to take my career & education further, and I am excited for future expeditions.
Last week was the last of our intense expedition on Coconut Island. I can say that it was the most special week for me, mainly because I learned how to use new instruments like the drones to better understand the populations of marine mammals which we studied, as (and not least importantly) to remember how much those 4 weeks that we have shared between our group of students, teachers, and assistants was an experience so enriching. I hope to be able to apply everything I've learned over the last few weeks in my work, and I want to say a big Mahalo to everyone who made this last month so special.
I'm going to be a little corny here and talk about the people doing the science, instead of the science.
While I may be a bit bias, there's something amazingly special (for lack of a less corny word) about the marine science community. Or maybe we're all just feeling the spirit of aloha? It's rare that I come into a completely foreign situation with a group of strangers and feel immediately at home. It often takes me months to feel the same level of comfort that I felt the first day I arrived to Moku o l'oe.
Each individual I've met here is quirky, kind, hilarious, intelligent, and inspiring. As this course is coming to a close, it's the memories of the people that will bring me the most nostalgia as I reminisce about this experience years from now. Thank you to everyone who made this course happen: the professors, the student assistants, the private donors, and all of the amazing postgraduates that I lived with for a month in Moku o l'oe!
Our time at Dolphin Quest and the module in general has been really thought provoking. While we had discussed our views on the various moral/ethical boundaries tiptoed by captive animal research, it was the overall direction of the research that enthralled me.
Almost every question we asked Lars during this week was met by an answer along the lines of "who knows". While normally that is an off-putting answer, in this context is was wonderful. It really felt like we were at some cliff edge of human knowledge when it came to these mammals, where so little detail is known. Using drones as we were doing in this situation really could be the difference between not knowing today, and starting to build a richer understanding of these animals down the line.
Rough seas ahead
One of the overall themes of this program I've enjoyed was how many problems we ran into at every step. The iterative process and figuring out solutions as a team felt so rewarding. It was phrased best by Einstein...
>If we knew what it was we were doing, it would not be called research, would it?
I'm definitely looking forward to finding more issues and overcoming them in applying new technologies to this realm of marine science.
Back when I lived in the small coastal town of Samana, in the Dominican Republic, between the ages of 12-15, every morning, for 3 months, I would grab my clipboard, data sheets, camera and GPS. I would head to the jetty to board one of the many whale-watching vessels full of excited tourists. From January to March hundreds of humpback whales come to Samana Bay, as it’s a key breeding and mating ground for North-Atlantic feeding groups. During this time, my parents allowed me to re-arrange my school hours (I was home schooled!) so that I could volunteer to collect data on humpback behavior for a local NGO gathering baseline data on these giant seasonal visitors. Thinking back, I realize that this was my first marine science experience and where I developed a strong connection with the ocean and the amazing animals that live it.
Fast forward to last week in Hawaii at the HIMB summer course. I had the chance to spend a week with marine mammal expert Dr. Lars Bejder and his student Fabien Vivier to learn about the new tech and methodologies they are developing to the study migrating cetacean across the world, including humpback whales. I was thrilled! On top of that that we got a crash course on flying drones, as these have proven to be really effective tools for finding, photographing and identifying whales. Lars and his team have been visiting breeding and feeding grounds of Right, Humpback and Grey whales across the world, and use drones to capture imagery. From the footage they are not only able to identify whales by name, but also take very precise body measurements, applying photogrammetry methods to estimate the volume and weight of each individual. From these data they have been able to keep close tabs on individuals, and mom and calves, and assess how taxing the long migratory journey is on their bodies. As a kid I was always fascinated by the fact humpbacks endured such long journeys from the high latitudes to their reproduction areas in the warm Caribbean waters and did not eat a thing during whole round trip -- as the warm water doesn’t t support the dense fisheries they rely on. I couldn’t have been more delighted than to have Fabian teach us how to measure and calculate these changes in weight from drone images, and work out that these whales lose between 25-40% of their body weight! On the other hand, we calculated that the weight of calves grew by 200-300% within a 2-4month period -- which they need to do in order to survive the journey to the cold waters where they will be weaned and their mothers can feed again to recover not just from the arduous journey but also from the added metabolic demand of producing nutrient-rich milk for their calves.
This week at HIMB's summer course "Applying Innovative Technologies in Marine Science," we are learning about remote sensing and how to use satellite images to measure coral reefs. Using satellite imagery from (free) software such as Google Earth and Planet, we can learn so much about how the earth is constantly changing on land and in the ocean. As a small example, I created this short video to highlight the devastating and historic flood my hometown and state is currently recovering from. The video shows three images taken on May 17th, 27th, and June 1st. I am amazed at the resolution of these images and the various questions we can answer using photographs taken from space!
Over the past few weeks we have been learning/playing/experimenting with 3D laser scanning and structure from motion (SfM) photogrammetry. Both of which are methods of scanning and creating a 3D model of physical object.
Laser scanning creates a 3D model of an object by rapidly recording the distance to an object with lasers from different angles. In a similar manner, SfM is an imaging technique for estimating three-dimensional structures from many two-dimensional image of the same object taken from different angles. Laser scanning an object is very quick, but the units are expensive (~20K) and cannot be used underwater. SfM photogrammetry is time consuming, but cheap (you can use a smartphone camera or a GoPro) and can be used in any environment you can take a camera.
Earlier post have touched on both methods and have highlighted how they can be used to address questions relating to coral reef structural complexity, volumetric changes in growth of corals, and changes in body condition of marine mammals undergoing long migrations.
But how do they compare on the same object in the lab?
It turns out if you get the technique right, and have a lot of patience you can get results from SfM photogrammetry that rival the expensive laser scanner.
You be the judge: Compare the images of the dolphin in this post with those taken by the laser scanner in earlier posts below.
This course has really opened my eyes to the multitude of methods that can be used to collect data, and how important it is to obtain data at different spatial and temporal scales to understand the full picture.
Last week, our group was in Dr. Liz Madin's module studying remote sensing with a particular focus on satellite imagery. At the end of the week we applied Google Earth and Planet imagery to our individual research interests to identify ways these tools can enhance our existing studies, and/or spark new interests for future work.
One of my research goals is to understand the seasonal variability of sediment transport systems (i.e. erosion and deposition) in Albany, Western Australia. My original plan was to collect drone imagery over the course of a few months (due to time constraints) and track the sediment over that time frame. However, through this remote sensing module, I realized that sediment patterns can be picked up through satellite imagery (as seen in the Google Earth photos below). This means I now have access to years worth of sediment changes, that will hopefully enhance and validate the trends I obtain from my drone videos! Woohoo thanks Liz!
Since May 13th I have been attending the “Applying Innovative Technologies in Marine Science" course at Coconut Island, Hawaii Institute of Marine Biology, University of Hawaii at Manoa. I cannot believe that the week that just started is our last one. Time flies!
Talking about flying… let me tell you about the amazing technology we were introduced to last week by Lars Bejder and Fabien Vivier from Marine Mammal Research Program University of Hawaii. We learnt the basics on unoccupied aerial systems (UAS) technology and about applying photogrammetry to produce, impressively accurate, body measurements on large cetaceans. The method includes reviewing whale aerial footage to capture the most adequate snapshot, followed by carefully applying dots around the whale contour to produce a series of body sections used to estimate their body volume. The pieces of the puzzle came suddenly together when we were able to detect quite dramatic differences in body volume suffered by these giants of the sea over time; in some cases because of the high energetic price paid to nurse their offspring, and in others as a consequence of the large migratory routes followed between feeding and breeding grounds.
Being research and conservation of marine mammals my field of work, I had been looking forwards to hear about it for some time and it did not disappoint me. On the contrary, what I learnt last week about using UAS to assess baleen whales body condition and population health via age-structure on dolphins got my head spinning on possible ways of implementing this innovative methodology to research I am conducting with the Ionian Dolphin Project (the project run in western Greece by Tethys Research Institute, the NGO I work with). I am eager to get back to my beloved Mediterranean and get started.
As a kid, the red & blue 3D glasses were always so exciting to wear, even when the 3D movies weren't that great. Now, in the HIMB Applied Innovative Technologies summer course, we found our childlike excitement through a different 3D technology - 3D modeling and printing. Group B (aka, the best group) had the opportunity to first laser scan a few dead corals (and a shark jaw) to develop 3D models of the specimens. We then took it a step further and pulled out a live coral cluster to laser scan and create a model. The next week, we were able to scan a small Hector's dolphin figurine, and the fluke of a live bottlenose dolphin at Dolphin Quest. Why do we want to do this? So much information can be gained when knowing basic information of corals, dolphins, and other marine fauna, such as volume, size, rugosity, weak points, probabilities of negative outcomes based on storms or temperature increase, and more. However, these animals live in the water and can only be out of the water for short periods of time. Also, most ways of calculating these metrics have been relative - and we are in the age of wanting absolute. That is the power of the 3D model. With this #cooltechHIMB2019 course, we are changing the way we look at studying marine life out of the water. What will be the next 3D modeling project at HIMB?
Out here on the islands, Aloha Fridays are a matter of celebration: an intention to take the week's most celebrated day with the ease and beauty that emerge from the spirit of aloha. On Moku o Le'a where scientists seem to be only outnumbered by the number of coconuts, we take time out of our Aloha Fridays to synthesize and reflect what we've learned.
Every week, our team has dove deeply into understanding and applying emerging technologies in marine research. Each module focuses on a technology poised to change how we collect data and make observations in the natural world, technologies like 3-D Scanning, underwater digital mapping, cutting edge satellite imagery, sensor development, and drone-based measurements of marine animals. On Aloha Friday we reserve the day to aggregate all we have learned, analyze the data we have collected, and discuss our conclusions about applying technology in science.
Synthesizing your results and making conclusions is the ultimate step in the scientific method, which establishes our process for developing knowledge from testing hypotheses with data. In a course like ours, Fridays are ripe for big ideas and engaging discussion about how these new sources of data influence our understanding of the natural world. This week, my colleagues and I spent hours collecting quantitative measurements about the growth and girth of marine mammals like whales and dolphins. On Friday, we sink our teeth into the exciting science of discussing how mother whales from different species all manage the process of calving and reproduction. While these creatures always seem huge to our naked eye, our careful measurements from UAV drones revealed dramatic changes in body size and condition over the process of raising a calf. In a month and a half, a lactating gray whale can transfer almost 15% of her total volume to a calf that grows almost 200% during that same time period. It's an astounding feat of mammalian biology, and taking the time to synthesize the results and reflect on what we learned trains us to understand what secrets of the natural world can be revealed through these new technologies.
To finish an incredible week in which we used different methods like photogrammetry and digital 3D scan to study the structure and allometry of the corals, we also had the opportunity to pilot an ROV - remotely operated vehicle (my first time)!
More than a third of our planet is covered by oceans, and they remain a big secret to us. An excellent tool for exploring unknown places and especially at great depths in the ocean is the use of ROV. It can be our eyes underwater and take us to unexplored places. It is equipped with video cameras and remotely operated on land or in a boat. The pilot sees where the ROV passes through the images generated and transmitted in real time on a monitor.
The use of new technologies, such as the case of Unmanned Aerial Systems (UAS), has recently been a great tool for assess the health of populations. Last week during the module "UAS photogrammetry for marine mammals", imparted by PhD Lars Bedjer and MsC Fabien Vivier, we learned to collect and process information by taking videos at differents heights of bottlenose dolphins in captivity using UAS´s. We made a comparison between the measures taken physically against the aerial measures and the preliminary results indicated the high precision and accuracy of this new technology, soon we will have more information about Fabien Vivier's PhD work. This module was definetely a great experiencen and learning!
There's nothing like the excitement of learning a new technology and finding innovative ways to integrate it into your own research interest.
My goal coming into this course on innovative technologies in marine science was to learn ways to apply these technologies to fisheries management, particularly for coastal communities to create self-sustaining management systems. Today was the last day of our remote sensing module with Dr. Liz Madin and something finally clicked. I was exploring the concept of "fishing the line" in which fishermen concentrate their fishing efforts on or near borders of no-take marine reserves to receive the benefits of the "spill-over effect". In other words, the area surrounding a marine reserve profits off of the over-production of fish within the reserve because, of course, fish are not stationary and don't acknowledge borders.
Using the Philippines as a case study, I began by looking at Google Earth to measure the abundance of artisanal fishing boats within 1 kilometer of marine reserve borders to see how the numbers changed throughout the years of available satellite imagery. I began to see a pattern in the number of boats increasing year after year. One would assume that this is the case because the fish catch must be increasing in these areas as well, right? In most cases it takes years after the establishment of a marine reserve to see a significant recovery of fish species so, while correlation is not causation, I began to think of the ways in which satellite imagery could identify determinants of marine reserve success. Could the number of fishing boats "fishing the line" of marine reserves be an indicator of ecosystem health? Could a free technology that is widely available and easy to use be an alternative to often time and money-consuming underwater surveys?
The answer is "I don't know", but I'm excited to explore these questions deeper and compare the satellite imagery with in-the-field reports of marine reserve success to understand if there is in fact a correlation. If so, this tool could be used by anyone, anywhere in the world to study the efficacy of marine reserves with limited funding, time, and technical skills.
Making your own underwater temperature logger might seem like a daunting task, but today each member of our team was able to build one from scratch! Using Arduino software (https://www.arduino.cc/) and various parts from Arduino including the Arduino nano, bme280 pressure temperature sensor and a nano data logging shield, we were able to program the logger to record temperature every few minutes. To waterproof the sensor, we embedded it in epoxy and made a custom rig out of PVC piping, using extra epoxy to seal the ends and we were able to successfully log temperature on the reef and keep our components dry! There are great instructions and example code online - making your own sensors are a fun and affordable way to support your research.
‘Photogrammetry, the breaker of chains’
Habitat complexity is important in coral reef ecosystems, since the biological infrastructures, holes, nooks, crannies and other features in reefs create shelter for associated flora and fauna. Although there are numerous metrics used to measure complexity in reefs, one of the most common is reef rugosity. Rugosity is a measure of surface roughness, and is commonly quantified using transect tapes and chains. The tape is laid at the reef in a relatively straight-line indicative of low complexity, while the chain traces all topographic highs and lows of the substrate (Image 1), with the ratio between the tape and the chain the rugosity value. However, using this method will only cover small areas and provide only a snapshot of reef complexity at one point in time.
Here at HIMB cool tech 2019, we’re learning photogrammetry (the science of making measurements from photographs) to build 3D models of reefs to measure rugosity and other metrics of complexity. The generated 3D models (Image 2 - didn't give it justice with my editing skills) closely resemble the surveyed reefs allowing the storage and tracking of reefs through time. Further, the spatial scale that we can survey is enormous, with 3D models of entire reefs possible. Moving forward in the advent of recent technology, we reef ecologist may slowly, but surely ‘break the chains’ of the chain-transect method and use more informative and accurate techniques to understand the importance of habitat complexity in coral reefs.
When working underwater, everything becomes at least five times more difficult. What seems straightforward on land is not such an easy task when you throw water into the mix! For example, we wanted to ground truth the measures of patch reefs taken from satellite imagery (see previous posts by Maria Isabel Gonçalves and Timothy Quimpo). It might seem like overkill, but we practiced unwinding transect tape on land, imitating exactly what we envisioned we would do underwater. When it came to the reality of the experiment, we started to understand what was and wasn't doable. Marine research gives us a great opportunity to use problem solving skills and innovation on the job to get the research done in time. Practice on land, make sure you have a plan before you jump into the water and be prepared for unexpected surprises, such as a tiger shark foiling your last reef measurement!
Having a background in coral reef ecology, I really enjoyed last week’s module in the J Madin Lab, learning about the new tech they are developing to study the structural complexities of reefs. For me the biggest highlight by far was learning how to create 3D photogrammetry maps of Hawaiian coral reefs from scratch. This involved going snorkeling in Kāneo`he Bay to survey one of its patch reef, “Reef 42”, which was a beautiful and impressive sight. For the first time, I witnessed a reef with 100% live coral cover. In the Caribbean, which is where I’m from and where I studied reefs, we get excited to see 20% live coral cover!
In the field the team and I took time-lapse images over a specific reef area. These images were then stitched together using a photogrammetry tool called Metashape, which uses digital elevation models to create a 3D “mesh” of the reef (see image). It was thrilling to see the 3D reef imagery come together, as it provides an invaluable opportunity to document in very great detail the structure of the reef, enabling us to carefully assess if and how it's changing over time. I was also delighted to learn how cost-effective this tool is, as the equipment is relatively affordable: compromising two GoPROs and a set of DIY gear mostly made up of PVC pipes, lead weights, clothesline wires, zip straps, and glue!
Last week I was introduced to some of the COOLEST technology used in coral reef research. We started the week using GoPros to photograph nearby patch reefs. From these photographs, we created an orthomosaic of the reef (basically a 3D picture collage) that can help us measure and identify coral species as well as look at changes in the reef over time. Then we used a new tool that is absolutely mind-blowing... a 3D LASER SCANNER?! I felt like Elon Musk at this point. Using this tool, we scanned individual corals and viewed their intricate details on the computer in 3D! We also built our own underwater sensors capable of measuring and logging temperature, pressure, and sounds. No big deal. And we drove an ROV (remotely operated vehicle). Safe to say we all geeked out! Overall the experience was 12/10. #cooltechHIMB2019
Our initial goal was to capture the morphometrics of some coral pieces using a 3D laser scanner. By the end of these attempts, with a bit of creativity, we had a figured out a process that gave us
painful acceptable results .
At the start, the best plan was to walk around a table in circles and slowly move the scanner to get good coverage. However, the stability was not great, and we frequently had incomplete models as we struggled to pace our angular rotation. We did improve after several iterations, but it left a lot to be desired. So when we went to the coral nursery, we had to mix it up.
Spin it right round
Although initially it sounded crazy, the new plan was to place a coral on a backless chair and then spin it around. This allowed the scanner to remain relatively static, minimising the risk of a bad scan. It took of a bit of getting used to, but eventually we were ending up with consistent scans. This approach seemed like a vast improvement, but eventually was replaced by something even better.
Return of the photogrammetry
After feeling like we had found a pretty solid solution to the scanning problem, it later transpired that going back to photogrammetry was even better. Using a similar spinning setup and a camera, Gus was able to get enough photos with a tight angular resolution to reconstruct the 3D model. Since this was based on higher resolution photos, he was able to not only see more fine structure of the coral, but then also implement real texture onto the model.
While there was not much exhaustive testing, or using a range of scanners with higher quality, it definitely looks like the photogrammetry approach was superior provided you had the computer power for the software. It also allows greater flexibility, being able to use photos from any source (quality may vary substantially), rather than needing the scanner. Something interesting to explore in more detail at a later date.
The costs of being a mommy whale. Many animals need to periodically migrate during their life span. Some do that to find food, some to reproduce. Several whales need to do both! That’s the case of Right, Humpback and Grey whales. They go to colder waters to feed and then travel to warmer areas to reproduce and have babies. But here comes a huge problem: these whales do not feed while migrating, reproducing and lactating. Can you imagine how much weight they lose before being able to feed again in cold waters? Although it is not possible to put whales on the scale, scientists are starting to have an idea of such costs by using technology! Scientists are using drones to capture photos and videos of the mom whales along time and with a computer software they can take measurements and then estimate volume and weight loss. Without this technological approach we would never know that some mom whales may lose more than ONE THIRD of their weight! Drones are helping us to see that, as it is for humans, having a baby is quite expensive for whales too!
For more information about whales and the use of technology to investigate them, check the link:
I remember being shown a Christmas tree worm on one of my first dives as a teenager - the dive instructor snapped his fingers in front of one and we all watched wide-eyed and delighted as the worm vanished in a flash, disappearing into its tube. I guess I've always had an eye out for them, and their (fanworm) sabellid cousins, since. Snorkeling around Reef 42 a few days ago, I was struck anew by just how rapid the curl-and-disappear response is in sabellids. When a worm is active, it has its feathery radiolar crown fully extended, which helps the animal respire and trap food in the plankton. If approached by a predator (or a swimmer with a camera, in this case), the radioles are rapidly withdrawn into the tube in which the worm lives. I had always assumed that this must be because of their ability to detect small changes in water motion nearby. But some reading around seems to suggest that sabellids also have eyes on their radioles that act as a "burglar alarm", visually detecting predators and initiating a rapid withdrawal response if a threat is detected. The eyes themselves don't seem particularly complex (more photoreception than true vision) and there seems to be a fair bit of uncertainty regarding just how they function and how they're wired to visual processing centres in the brain, but I was quite surprised to learn that these worms even had eyes. Seems a bit risky, too, to grow eyes on such fragile appendages. But turns out the entire radiolar crown can be regenerated within a few days or weeks if its damaged, with all the visual machinery in tow - what I would assume is a pretty complex, somewhat costly, endeavor. This makes me wonder what's eating these fanworms in the first place - but maybe that's a wormhole for another day.
Sciencing by the couch. Here at the 2019 HIMB summer program, we’re learning that ecological data can be collected while sitting at your living room, or office couch. For this exercise, we measured the perimeter and area of patch reefs along Kaneohe Bay using satellite imagery from Google Earth and Planet. The imagery allowed us to survey a large area that would otherwise have taken a day or two of fieldwork. We were also able to ‘go back in time’ to see potential changes at each patch temporally. Our ability to now survey reefs at large spatial and temporal scales will open new avenues of research on coral reefs while sitting comfortably at home.
Today, the group that is working in the Remote Sensing Imagery module, performed a field trip to measure coral patch reefs. This week we are learning how aerial images can be used in a wide variety of studies, such as measuring ecosystems of interest, counting animals found in areas of difficult access such as polar bears, whales, and using this type of images to aid in the management of marine protected areas.
With the data collected today, we will compare measurements made in the field with measurements made from aerial images at the lab. We still had several exciting moments during the field and were able to see several turtles and sharks. Great day!
Do you know that technology can help to the conservation of endangered species? That was the topic of a great lecture of Charles Littnan during the first week of the course. In his talk we could learn how innovative approaches may help to understand Hawaiian Monk Seal biology and to generate information for its conservation. Hawaiian Monk Seals are only found in the Island of Hawaii and its population has been decreasing as a result of both natural (shark predation, diseases) and human threat (accidental bycatch in fishery activity, habitat destruction). As they are spread across Hawaiian Islands, it is difficult to monitoring them and even to know how they live. That’s why technology can be a good ally. Researchers are using drone images and bio-logging tags to track their distribution and to monitor their population and behavior. Additionally, video cameras attached to them are helping to understand Monk Seals habits and to show that they are not a threat to the coral reef environment and to fisheries. A great thing is that not only scientists can use technology to help in the conservation of Monk Seals! Everyone can do that using social media (Instagram, Facebook, Twitter) to post photos of Monk Seals tagged with #MonkSeal, so scientists can gather these photos and generate scientific data with them. You can find more information about Monk Seals and their conservation at:
We are running a summer course at the Hawaii Institute of Marine Biology in Kaneohe Bay. We are using a range of technologies to explore marine mammals biology through to remotely sensed and three-dimensional structure of corals and coral reef. There are 12 participants from around the world, and we will track our summer expedition here.
Contribute to this expedition
Thank You for Your Contribution!