Forest Flows
The "kitchen sink" approach to forest hydrology
"The Forest Flows Programme is delivering the science for evidence-based decision making for managing water use, which the forest industry is investing in"
Paul Adams - CEO Forest Growers Research
Black box
The five-year Forest Flows research programme wrapped up in June 2024. Scion senior scientist and programme leader Dean Meason “chucked the kitchen sink” at hydrology research in New Zealand to provide a better idea of exactly what is happening to the water in our forest soils.
Described as a "black box", there have been few advancements in hydrology research over the last 40 years. But the complex processes of how water is distributed, used and circulated in forested catchments is now clearer.
BIG DATA
This ambitious programme set out to combine a slew of different measurements taken at different scales all the way from space to one metre below ground. Everything from the air, rainfall, tree growth, the age of water, soil composition and more were recorded and validated. The first step was establishing the sensor networks and how the researchers could access this huge influx of data.
More than 390 million measurements were captured over the programme with 1717 sensors established across 9 study sites in New Zealand. Mapping this data with remotely sensed measurements (NASA satellite imagery and multispectral radar imagery from a plane) was central to the research.
Data alone doesn't solve the problem.
Five primary sites
Between 2019 to 2021, Forest Flows selected sites and established research plots at various locations around New Zealand. Working with forestry industry partners, field crews set up 25 research plots at each of the five primary sites.
Locations were chosen to cover a rainfall gradient.
TE hiku forest
Forest catchment on sand dunes
1200mm annual rainfall (moderate)
We found a difference in soil water dynamics between younger sand and older marine sand deposits. Te Hiku had herds of wild horses regularly grazing through the catchment. Although this forest was on sand dunes, there were also lots of wetlands present.
Mahurangi Forest
Unexpected weather events
1600mm annual rainfall (moderate)
Just north of Auckland, this catchment was hit hard by extreme rainfall events in 2023. Over the space of five weeks, Cyclone Hale, the Auckland Anniversary weekend flooding then Cyclone Gabrielle blew through.
Despite all the rain, the catchment was able to hold onto a large amount (approximately 60%) of the rainfall.
titoki Forest
Radiata pine act as a nursery species
1200mm annual rainfall (moderate)
The radiata pine at this catchment provides shelter and protects the establishment and growth of native plants on land that was in pasture for more than 100 years.
Water would pop out from underground to supply the stream all year round.
Mawhera Forest
Steep and deep
3000mm annual rainfall (high)
This forest is situated over extensive gold mining works at the lower part of the catchment.
The stream channels at Mawhera are steep and deep - at some places they reached a depth of 10m. This shows how much water flows through West Coast (high rainfall) catchments.
Ashley forest
Disappearing streams
800mm annual rainfall (low)
The stream in Ashley Forest is described as ephemeral - it would disappear underground and reappear further downstream.
The groundwater measurements from this forest provided crucial data for understanding water movement and is important for water release from the catchment.
Additional sites
Beyond these key locations, four other forests (Balmoral, Rewanui, Riverhead and Puruki) were also monitored for additional data e.g. native species were investigated at the Rewanui site.
cutting edge technologies
The success of Forest Flows is underpinned by the development of cutting-edge technological solutions.
- Novel datalogger equipment and wireless networks
- Innovative methods to measure soil moisture from the air
- A computer-generated forest digital twin
- Unprecedented hydrology measurements
- A pipeline to handle big data volumes
- Machine learning and artificial intelligence tools for data analysis
Not only did we create and develop cutting edge science to capture what was happening to water movement in New Zealand Forests, we were able to answer pressing questions for local communities.
The sheer volume of incoming data presented a problem, and in a New Zealand first, Scion developed the Kafka Big Data Pipeline that seamlessly cleans, summarises and securely stores big data arriving from the forest sensor networks in near real-time. This cloud storage capability made it simple for national and international collaborators to access and analyse the data.
Deep Learning
Forest Flows collaborated with the MBIE Time-Evolving Data Science/Artificial Intelligence for Advanced Open Environmental Science programme led by the University of Waikato’s Artificial Intelligence Institute. This collaboration enabled valuable synergies to both, with Forest Flows providing complex environmental data streaming in near real-time.
A range of new deep learning methods were developed:
- predicting temporal/spatial variability of tree growth/water use within and between sites with a Deep Learning Neural Network
- forecasting tree growth and water stress with AI
- directly forecasting from data streaming from the forest
These methods are suitable for other ecosystems with complex environmental data.
Common misperceptions
Last year, these world-leading results were shared at the annual Forest Growers Research Conference in Dunedin, New Zealand. Common myths about radiata pine forestry have prevailed in New Zealand and Dean was able to counteract these with the new results fresh from Forest Flows.
Myth: Forest canopies intercept 42% of water
Results showed that canopy interception was lower than expected and ranged from 9% to 27%, with 9% for the lowest annual rainfall site at Ashley Forest in Canterbury. Rainfall still moves through the canopy and enters the soil during low rain events.
myth: trees use all the rainfall during summer
We found that rainfall entered the soil during low rainfall events. Water regularly drained from the root zone during low and high rainfall events.
Annual tree water use on average was 41% (range 16%-51%).
Rainfall amount doesn't directly relate to percentage of water used by the trees
Annual rainfall is lowest in Ashley forest to highest in Mawhera
The amount of water leaving as streamwater is linked to rainfall
But, more water left the catchments as groundwater than as streamwater for 3 of the 5 sites.
Myth: Indigenous species use a lot less water than exotics
We found that some native tree species used proportionally similar rates of water as exotic tree species.
Analyse and simulate
Forest Flows used a suite of statistical, empirical, modelling and AI methods. No one method was able to answer all the objectives of Forest Flows, rather they were addressed using the strengths of each approach.
The Forest Flows hydrological model links forests to the stream and provides accurate predictions of tree water use and stream water flux. This model is the combination of two models:
- CABALA-W: Upgrade of the well-regarded process-based forest model CABALA
- Pauwels hillslope model
The Forest Flows Hydrological Model produces realistic predictions both for stream baseflow (between rainfall events), stormflow (during rainfall events) and the rate streamflow decreases after rain. CABALA-W provides accurate predictions of canopy interception, tree water use, and soil water movement and storage.
Scion, NIWA, Moinhos de Vento Agroecology Centre and CSIRO developed the Forest Flows Hydrological Model.
Myth: planted forests pollute waterways
There are concerns that planted forests pollute waterways with large nutrient runoff. Water quality monitoring provides a snapshot in time, typically sampled once monthly. But how does water quality change throughout the month?
We measured hourly stream nitrate levels, directly inside planted forests.
What we found was that stream nitrate concentrations measured in streams in planted forests are low. The changes in stream nitrate concentrations during rainfall events show that mean stream nitrate concentration should be weighted by the volume of stream water. Stream nitrate concentrations weighted by streamflow volume are lower than other primary sector land uses. Continuous nitrate monitoring has provided new insights that would not be possible with once-a-month water sampling.
Forest flows was a constructive engagement for various community groups including Federated Farmers… Thanks to all the scientists who helped break down the findings in an informative and interactive manner”
Research doesn't exist in a vacuum - information needs to be shared with local communities, end-users and stakeholders in a format they can use.
Reaching communities
Building trust in research outputs from end users is essential for creating a meaningful impact. This was nurtured through transparency and clarity about what our research set out to do. Additionally, it involves understanding the various challenges that stakeholders faced in their daily management and operations.
Over the programme our scientists ran 37 workshops, wananga, hui and other engagement activities to talk with key stakeholders, even during the COVID pandemic. They were upfront about what the research could and could not deliver. As a result, when the programme's findings became available, stakeholders were eager to participate and even helped identify other parties who could benefit from the results.
Ngāi Tahu Forestry are pleased to partner with Scion on Forest Flows to better understand how plantation forestry impacts the flow and quality of water both within and beyond their forest estate."
"This project reflects the increasing importance being placed on the protection of our natural resources and ecosystems and Ngāi Tahu envisage that the information obtained by this programme has the potential to enhance and strengthen beliefs and values with respect to the protection of the land in perpetuity."
Dean is really proud of the engagement and trust the team built. "Engaging early, sharing information and the two-way communication; it helped us define the questions that Forest flows needed to answer for local communities and what information stakeholders wanted."
The most recently published paper from Forest Flows is Trees and water: A survey of the perception and decisions of landowners in New Zealand.
Long-term Impacts
The overall benefit for Aotearoa New Zealand is to provide accurate data about tree water use, catchment water storage and release in planted forests.
"Forest flows has addressed a critical knowledge gap. Evidence generated by the Forest Flows programme has been used during the Otago Regional Council’s (ORC’s) freshwater plan development process to counter proposed policies with catastrophic outcomes for the forestry sector.
"The evidence provided by Forest Flows was able to correct long held myths on forest hydrology and, whilst the process is on hold for now, the setbacks proposed by ORC will be significantly reduced."
The results are helping industry with their decision-making, workshops with iwi and community groups have shown the high interest and willingness to learn about water in our forests.
Forest Flows has provided information to stakeholders, decision makers and the public for managing existing planted forests and water, as well as those deciding to plant new forests.
We specifically designed forest hydrology models that can be used to test planted forests' impact on surface and groundwater under current and future climates and allows the quantification of water yield.
Forest Flows in Numbers
1717 sensors
across New Zealand, 360 000 observations daily
154 plots
18 different types of ground sensors at the plots
87 People
doing the research from 23 organisations
31,000 km
total distance driven for fieldwork and maintenance
What's next
One remaining myth to tackle is that "very little water leaves exotic forest catchments in summer".
Data analysis on water release from forests is ongoing. In a previous study at Purukohukohu paired catchment, we showed that more streamflow left the catchment from radiata pine than those in pasture.
Expanding from forested catchments - looking at the whole landscape rather than a single forest you get a much better picture of water movement.
We are continuing to improve how we use remotely-sensed data from satellites, to give new insights of the hydrological cycle under different landuses and across the landscape.
NASA project built on Forest Flows
In 2024, Scion collaborated with NASA on a feasibility project built on the immense data networks and methodologies developed during Forest Flows. This successful feasibility project opened up further opportunities to collaborate with NASA in 2025 and beyond.
The NASA-Scion feasibility study looked at developing new machine learning prediction methods to measure water flux above and below ground across forests and different land uses in the Aupōuri Penninusla, and working with the Te Hiku community - including the Te Hiku Forest Collective (Te Rūnanga Nui o Te Aupōuri, Te Manawa o Ngāti Kuri, Te Rūnanga Nui o Ngāi Takoto, and Te Rūnanga Nui o Te Rarawa Trusts).
The collection of data at Forest Flows sites is continuing with high interest from the Forestry industry. This extended monitoring is supported by funds from Forest Growers Levy Trust and Otago foresters.
“Spreading the word further, that’s the next challenge.”
This research was funded by Ministry of Business, Innovation and Employment Endeavour Fund. Extended monitoring and data collection at Forest Flows sites is supported by FGLT and industry.
Related stories
When the rain pours, where does the water go?
Extreme weather events are becoming more commonplace.
MaiMai forest, home of hydrology
In June 2023, MaiMai catchment celebrated 50 years of hydrology research in New Zealand.
Scion and NASA collaborate
Two research projects to investigate wildfire risks, and water flows were funded in 2024.
Find all the scientific articles published from this programme here. This will be updated as more results are published
Collaborators
Many thanks to all our research partners, collaborators and people who made this programme possible.
|
Candleford |
||
|
Whitegum Forest Natural Resource (Australia) |
Keen to know more? Visit scionresearch.com/forestflows
Built with Shorthand