Honey Estarija is a self-taught unmanned aerial vehicle or drone pilot specialising in detecting myrtle rust. Photo / Supplied
Honey Estarija is a self-taught unmanned aerial vehicle or drone pilot specialising in detecting myrtle rust. Photo / Supplied
Accurate measurement of carbon surrounding trees can now be made using laser scanners, says Sadeepa Jayathunga.
The remote sensing and forest management spatial specialist is one of six leading Scion scientists working on technology for forest monitoring and climate resilience planning. She said she felt confident about presenting her work to Nasa and other industry experts during this week’s ForestSAT 2024 conference in Rotorua.
One part of the Scion Transforming Tree Phenotyping and Forest Insights team, her research was funded through the Ministry of Business, Innovation and Employment (MBIE).
Already recognised for providing local forestry insights through their work, particularly in the Kaingaroa, Timberlands and Kinleith forests, the scientists focused on radiata pine trees during their pilot run.
Four of the seven scientists presenting at the conference said they wanted to continue developing their databases to include New Zealand’s native flora such as pōhutakawa.
To measure carbon around pine trees, technical challenges were solved without a “supercomputer” — something scientists initially said couldn’t be done.
But Jayathunga, who was self-taught, managed to create the necessary “pipeline” without one, determined to “use what we had” and push herself and her computer to its limits.
Overcoming cost barriers was a “proud moment” and a breakthrough for climate change management in the future, she said.
Jayathunga said her challenge was to reduce data processing times, and it required processing pipelines that didn’t exist yet.
Eventually calibrating a “pipeline” or system that could process the free regional Lidar data provided by regional councils, Jayathunga said she was pleased to make use of the information.
“Writing all these scripts and developing the tools I didn’t even know I would be able to do it.
“It was a really hard time, working really late nights and trying to figure out why your script is not working, why it’s spitting out errors.
“We were really struggling to process that data to get meaningful insights because we have limited computer resources.
“I started developing all the data processing pipelines from scratch because we didn’t have anything to process big data sets and started streamlining the pipelines.”
Her development data processing used to take months, but Jayathunga’s pipeline cut the processing times down to days.
Scion principal scientist Michael Watts is proud of the work his team have done on the Forest Insights prototype developed by his team. Photo / Supplied
Michael Watt, principal scientist at Scion, said the Tree Phenotyping programme focused on the Bay of Plenty. The Forest Insights programme was a national-scale project that produced “progressive” and “cutting edge” results.
Progress made by his team allowed them to move into precision forestry, venturing into hyperspectral and thermal data, tech developed could successfully pre-visually detect myrtle rust in pōhutukawa trees for the first time, Watt said.
“I think we’re at the cutting edge and yeah, I’m pretty proud of what we’ve done.
“Sadeepa’s work’s really shown for the first time that you can predict these tree dimensions at the individual tree level, which gives you a lot more information than you would normally get.
“Changes in carbon are very important for mitigating the rate of climate change because obviously, you don’t want to lose carbon that’s stored in trees, they store a huge amount of carbon,” Watt said.
Information collected would be used to “accurately characterise the volume and carbon in those forests at the tree level”, Watt said.
Honey Estarija in the Scion lab.
Arriving from the South Island in 2021 as an intern, Melanie Palmer is now a data analyst at Scion.
Her role on the Forest Insights team was to specialise in remote sensing and geospatial and spatial analysis.
As a visual person, she enjoyed working with aerial images, maps and using AI as a scalability tool.
“I don’t see AI as something that replaces people, it’s something that we use alongside productivity tools.
“For Forest Insights, when we’re producing this map over an entire country, doing that by hand as an analyst is rather painstaking; to be able to get computers to do that for us is really powerful.
“Suddenly, it’s like, oh man, I didn’t even know there was some forest over in this far-off corner,” Palmer said.
A sample image of spatial analysis data from the Forest Insights project.
Honey Estarija started at Scion as a student after studying resource management at Toi Ōhomai. She taught herself to be an unmanned aerial vehicle (drone) pilot.
Her role on the Tree Phenotyping team was to work with thermal imagery to detect tree health, diseases, tree growth and their form.
As the project leader for identifying myrtle rust in pōhutukawa trees, she said she would love to develop the work outside a laboratory.
“That’s actually my favourite tree, a culturally significant tree in New Zealand, and it is highly vulnerable to myrtle rust,” Estarija said.
Esatrija said she was proud to develop biosecurity measures ready for international and commercial markets as part of the project.
The potential to take the test run of myrtle rust in pōhutukawa trees outside the lab and into local forests was something she looked forward to after the conference.
Registrations for the conference were open and Scion said they welcomed residents interested in forest health and management to attend.
Aleyna Martinez is a multimedia journalist based in the Bay of Plenty. She moved to the region in 2024 and has previously reported in Wairarapa and at Pacific Media Network.
