Information Needs for Land Stewardship in Fire-adapted Landscapes

Meeting land managers’ information needs in today’s fire-adapted landscape

By Sean Reilly, MPhil Environmental Change and Management (2020)

Pepperwood Preserve after the October 2019 fire. Pepperwood Preserve, Sonoma County, California Photo Credit: Gerald and Buff Corsi

As every Californian knows, fires play a significant role in our landscape. Our forests and other natural systems evolved under a cycle of intermittent, low-intensity fires. These fires clear out the forest understory but leave the canopy largely undisturbed. However, since the early 20th century, aggressive fire suppression tactics statewide have reduced the frequency of fires. Unfortunately, reduced fire frequency results in more fuel buildup between events. Therefore, when a fire does burn, it burns with the uncontrollable intensity that we have seen on the news and experienced in our communities in recent years. Unless action is taken to change course, this phenomenon will worsen. As a result of climate change, the average area burned per year in California by uncontrolled fires is projected to increase considerably by the end of the century (Westerling, 2018).

Despite the extensive focus this topic has received in recent years, California’s most recent Climate Change Assessment Report identified critical knowledge gaps around the ability of land managers to implement mitigation measures due to the inaccessibility of information relating to forest health and resilience. In order to target efforts that effectively minimize fire hazards, land managers need access to site-specific information on fire hazards, such as forest structure and fuel load, that increase an area’s vulnerability to wildfire.

Sean received a Fulbright Scholarship to conduct a community-based participatory research project with the Mandingalbay Yidinji aboriginal people of Australia to manage the pond apple, a formidable bioinvader valued by the community as a food source.

In 2017, I spent a year in Australia mapping a population of pond apple, an invasive tree, for the Djunbunji, an Aboriginal ranger group. They wanted to take action against the species, but their previous efforts had been unsuccessful since they did not know the extent of the tree’s distribution and it had, therefore, been able to re-spread from surviving pockets they had previously missed. In situations such as this, where information is difficult to acquire, land management objectives often become unattainable without the input of significant time and resources. In an effort to improve their future chances of success, I spent months crawling around in the mangroves trying to find isolated populations of the pond apple. I was attempting to be as systematic as possible, even though I was lost and afraid a snake (or possibly a crocodile) was hiding behind the next tree.

As the months dragged on, I grew frustrated with the sheer effort and time required for this manual sampling. My frustration increased in September when I began to notice the leaves of this species turning yellow while all other nearby vegetation remained solidly green. When I returned to the United States, I decided to do some additional research using Google Earth to check my suspicions and, as I had suspected, the populations I had spent months diligently mapping on foot could be clearly seen as bright yellow patches in the otherwise green sea of the mangroves! The Google Earth map was certainly not as precise or complete as the one I had constructed manually but would have been more than sufficient for the rangers’ needs. At the time, I was familiar with the capabilities of remote sensing but had never seen so concretely what an advantage the technique could provide to land managers.

The reality is that land management, whether invasive species removal in Australia or fire hazard mitigation in Sonoma County, is largely a triage operation since almost every operation is plagued by insufficient time, money and resources. Management actions often boil down to “second best” approaches, whereby lofty objectives are balanced with real-world constraints to develop plans that are realistically achievable. Given this reality, having perfect data, in this case about the tree population, does not actually improve outcomes due to the additional time and resources required to obtain it. A coarser map would have fit the rangers’ needs but, with my solo ground method, there was no less time intensive approach I could take.

This drone is mounted with a multispectral camera. By taking lots of photographs with a high amount of overlap, we are able to actually create 3D models of the ground (or forest) we are flying over. This method is called Structure-from-Motion.

Nonetheless, environmental information gathering is rarely as straightforward or analytically simple as the yellowing of a single tree. This is especially true for mapping fire hazards in the varied ecology of California and, particularly, Sonoma County. Both the scale and the complexity of the problem seriously challenge any method for information gathering. Currently, new remote sensing techniques and technologies are being developed for assessing forest health over large areas. The three main technologies are airborne LiDAR (pulsed lasers for distance measurements), drones and satellites. They could be more useful in practical management situations than current field methods since they have the potential to be more cost effective, easier to utilize and cover larger spatial extents.

The Command Station is hooked up to a drone (pictured left) on which there is mounted a multispectral camera. The camera takes lots of photographs with considerable image overlap. The technology works by finding the same features in overlapping images. By comparing how their relative positions vary between the images, it is possible to determine their locations in three-dimensional space. The goal is to compare the results we get from this technique to the airborne and ground-based LiDAR data to determine how close we can get with this method since it is cheaper, easier and faster.

However, these tools will only be truly useful if they meet the specific operational needs of land managers working on the front lines to improve forest resilience against future fires. To further confound matters, different groups and individuals engaged with land management have very diverse informational needs and capacities. It is for this reason that I am currently conducting a study to gather perspectives from land managers on the role of information in their decision-making. I am conducting this study as part of my Master’s degree at the Environmental Change Institute at the University of Oxford. The objective of this study is to identify the common uses and requirements for information gathering techniques to improve their utility in real world applications.

I am employing a technique known as Fuzzy Cognitive Mapping, whereby interview participants identify key components of their decision-making process around managing fire hazards and the relationships between them. The goal is to understand how different sources of information flow through this system and whether they enhance or hinder decision-making. I will compare maps that are created by participants from all backgrounds of land management, from professional foresters to community landowner groups. Through this comparison I can identify the core attributes that should be prioritized in order to maximize their practical usefulness. I am collaborating with Dr. Lisa Patrick Bentley and Dr. Matt Clark at Sonoma State University to understand the ecological relevancy and implications of this information. Hopefully, using our combined efforts, understanding fire hazards will become easier and more cost-effective and will, therefore, allow land managers to effectively manage wildfires to prevent the uncontrolled events we have witnessed in recent years.


About the author: Sean Reilly graduated from Santa Clara University in 2016 with a BS in Environmental Science and Biology. While at SCU, he conducted research on atmospheric river storm systems and served as the president of an environmental group for three years. In 2017, Sean received a Fulbright Scholarship to conduct a community-based participatory research project with the Mandingalbay Yidinji aboriginal people of Australia to manage the pond apple, a formidable bioinvader valued by the community as a food source. As a Rhodes Scholar, Sean is pursuing an MPhil in Environmental Change and Management on forest fire risk mitigation in California.

If you have any questions about the study or would like to take part in an interview, please email me at

Works Cited: Westerling, A., 2018. Wildfire simulations for California’s Fourth Climate Change Assessment: Projecting changes in extreme wildfire events with a warming climate (No. CCCA4- CEC-2018– 014). California’s Fourth Climate Change Assessment, California Energy Commission.

Post a comment

Translate »