The world’s leading publication for data science, AI, and ML professionals.

A Closer Look at the Biomass Burned in California’s Wildfires

Background

Data for Change

We compared tree species and biomass data from UC Santa Cruz’s ForestGEO site to wildfires like the one just northwest of Santa Cruz proper and here’s what we found.

Last week, the state of California set a new record. Over the course of the year, more than two million acres of forest have gone up in flames. That’s about five times the size of New York City, and then some.

As if all the other events in 2020 weren’t wreaking enough havoc.

Wildfires of the west are not new. In fact, the California Department of Forestry and Fire Protection has been tracking them since 1932. The fires have heightened, particularly in the past 20 years, in part because of climate change induced drought, human oblivion and extreme winds.

With wildfires continually on the rise, it’s crucial to acknowledge and understand the impacts of burning vegetation on the environment. Kilograms upon kilograms of smoldering biomass can substantially increase the amount of net carbon released into the atmosphere. Put simply, this uptick in carbon levels in turn heats the earth and contributes to global warming.

When Emma Scott and I discovered the drastic sway wildfires can have on Climate Change, we were driven to learn more about the potential effects of burning biomass on the most vulnerable ecological communities. To begin, we decided to estimate the biomass lost in the California wildfires.

Part 1: Analyzing California Tree Censuses

For us, a key component in estimating burned biomass was to survey the state’s forest landscape.

California is vast. As students, we don’t have the resources to travel along the West Coast and record forestry data. We did, however, have access to the Forest Global Earth Observatory’s (ForestGEO) tree censuses. ForestGEO is a network of scientists worldwide who, across 72 observational forest sites, aim to advance research of the Earth’s ecological structures. In order to accurately track tree growth and maintain the observational region, most ForestGEO sites regularly produce tree censuses detailing each tree’s species, coordinates, diameter and stem counts.

Given that many of the wildfires in California are ravaging through metropolitan areas, we chose to narrow our focus on ForestGEO’s University of California Santa Cruz site.

The Southern Bay Area and a pinned location of the UC Santa Cruz ForestGEO site (image by UC Santa Cruz ForestGEO researchers)
The Southern Bay Area and a pinned location of the UC Santa Cruz ForestGEO site (image by UC Santa Cruz ForestGEO researchers)

Just some quick facts about the UC Santa Cruz ForestGEO site:

  • The primary tree families are coastal evergreens and redwoods
  • The site’s area is just under 40 acres (roughly 16 football fields)
  • The site contains 31 tree species
  • The site’s total tree count is 8,230 (as of 2013)

Beyond noting some of the quick stats and facts, we were curious to take a look at the breakdown of tree species within the site.

A rough map of the 15 most common tree species in the UCSC ForestGEO site, with species mapped to color (image by author)
A rough map of the 15 most common tree species in the UCSC ForestGEO site, with species mapped to color (image by author)

Based on coordinate locations of individual trees, we developed a plot of the 15 most common species within the UC Santa Cruz ForestGEO site. Douglas fir (Pseudotsuga menziesii), California live oak (Quercus agrifolia) and tanoak (Notholithocarpus densiflorus) were among the greatest in frequency.

A chart of the most common tree species in the UCSC ForestGEO site by individual tree count (image by author)
A chart of the most common tree species in the UCSC ForestGEO site by individual tree count (image by author)

Taking our analysis a step further, we used the allodb package in R to find the number of trees per species and their cumulative above ground biomass contribution to the forest.

Considering the average height of a Douglas fir rivals a 10 story building and they are the most common tree at the UC Santa Cruz ForestGeo site, it was no surprise that they accounted for a large percentage of the forest’s above ground biomass.

A chart of the the UCSC ForestGEO site's tree species with the greatest cumulative biomass (image by author)
A chart of the the UCSC ForestGEO site’s tree species with the greatest cumulative biomass (image by author)

As a result of having a high tree count, some species, like Douglas fir and California live oak, make up a high percentage of the UC Santa Cruz site’s biomass. While the Ponderosa pine (Pinus ponderosa) and Big Leaf Maple (Acer macrophyllum) are not among the most populous species, because of their sheer size they still significantly contribute the forest’s above ground biomass.

Part 2: Identifying a Comparable Wildfire

After locating and evaluating the UC Santa Cruz ForestGEO site, we found a burned area of similar ecological composition. The CZU Lightning Wildfire complex was designated as a sister site because it closely resembled the UC Santa Cruz site’s elevation and tree species breakdown.

An outline of the area burned by the CZU Lightning Wildfire, just northwest of Santa Cruz (image by Piedmont Exedra)
An outline of the area burned by the CZU Lightning Wildfire, just northwest of Santa Cruz (image by Piedmont Exedra)

The CZU Lightning Wildfire:

  • Was caused by lightning strikes in August 2020
  • Became 100% contained as of mid September
  • Stayed active for 37 days
  • Affected 2 counties: Santa Cruz and San Mateo
  • Destroyed 1,490 structures
  • Burned 186,509 acres

The area burned by the CZU Lightning Wildfire was geographically nearby the UC Santa Cruz site and consisted of a number of state reserves including Big Basin Redwoods State Park, Butano State Park and West Waddell Creek State Wilderness. While unofficial species lists are available for these forests, there are no existing resources that determine the exact forest composition.

Research by ecologist Roy W. Martin shows that the __ UC Santa Cruz ForestGEO site and CZU Lightning Wildfire do have plant communities in common, namely evergreens and redwoods. It is these commonalities between these sites that allow us to consider them sister sites.

Part 3: Modeling and Estimating

Since the UC Santa Cruz site and CZU Lightning Wildfire area have a common forest composition, we can use the biomass calculation of the ForestGEO site to inform a biomass approximation at the site burned by the CZU Lightning Wildfire. The area of the UC Santa Cruz ForestGEO site is 39.54 acres and contains a total of 4,102,997 kilograms of biomass. Similarly, the area burned by the CZU Lightning Wildfire is 86,509 acres. To estimate the biomass consumed by the CZU Lightning Wildfire, we can solve the following proportion:

A calculation of the approximate biomass burned by the CZU Lightning Wildfire (image by author)
A calculation of the approximate biomass burned by the CZU Lightning Wildfire (image by author)

Solving for the unknown yields almost 9 billion kilograms of biomass burned (8,976,888,403.4 kilograms, to be exact).

An olympic swimming pool holds 1 million kilograms in mass. The magnitude we’re talking about it 9,000 olympic swimming pools of trees. Vanished in the course of a single fire.

This is just a first pass at gauging the biomass burned by the CZU Lightning Wildfire. To make estimates more precise, we hope to conduct some detailed spatial and elevation analysis.

Just by dropping into forests on Google Maps, we can tell that there is a relationship between tree species and elevation. Logically this checks out as well. When hiking hills it’s easy to observe that tropical-esque species tend to appear at the base, and, above the tree line, the environment quickly turns alpine. We’re hoping this observation can shed light onto the percentage breakdown of various tree species at CZU Lighting Wildfire site.

Preliminary research tells us that Douglas firs inhabit lower elevations while knobcone pines occupy higher ones. If, for instance, the forests impacted by the CZU Lightning Wildfire was about 90% low elevation and 10% high elevation, we may be able to suggest the site was 90% populated with Douglas firs and 10% populated with knobcone pines.

Of course, this rudimentary example is only a rough estimate.

Reflecting on the Research Experience

This project didn’t strictly follow material outlined in the textbooks reviewed during the course (Forecasting: Principles and Practices and Ecological Forecasting). Consequently, out project was fairly open-ended. We weren’t given instruction on how to proceed with research. Instead, Emma and I looked to each other for inspiration and Professor Kim for mentorship.

Although it was challenging to determine our path while conducting research, this project provided me with some insight about developing a Data Science project. Experiential learning has proven to be one of the most valuable parts of my college experience. I look forward to using these skills in the workplace in future endeavors.

References

Pierre-louis, Kendra, and John Schwartz. "Why Does California Have So Many Wildfires?" The New York Times, 21 Aug. 2020, nytimes.com/article/why-does-california-have-wildfires.html.

Kolomatsky, Michael. "How Big Is an Acre, Anyway?" The New York Times, The New York Times, 26 July 2018, nytimes.com/2018/07/26/realestate/how-big-is-an-acre-anyway.html.

Salo, Jackie. "2020 Events so Far: Yep, These Major World Events All Happened This Year." New York Post, 10 Nov. 2020, nypost.com/list/major-2020-events/.

"Top 20 California Wildfires" California Department of Forestry and Fire Protection, 3 Nov. 2020, fire.ca.gov/media/11416/top20_acres.pdf.

Williams, A. Park, et al. "Observed Impacts of Anthropogenic Climate Change on Wildfire in California." AGU Journals, John Wiley & Sons, Ltd, 4 Aug. 2019, agupubs.onlinelibrary.wiley.com/doi/full/10.1029/2019EF001210.

"People Are behind Costly, Increasing Risk of Wildfire to Millions of Homes." ScienceDaily, 10 Sept. 2020, sciencedaily.com/releases/2020/09/200910130410.htm#:~:text=People%20are%20starting%20almost%20all,-threatening%20wildfires,%20researchers%20report.

Freedman, Andrew. "’Particularly Dangerous’ Wildfire Threat to Unfold in Southern California with 75 Mph Santa Ana Winds." The Washington Post, 2 Dec. 2020, washingtonpost.com/weather/2020/12/02/santa-ana-southern-california-wildfire/#:~:text=The%202020%20wildfire%20season%20in,to%20evacuate%20on%20short%20notice.

Dunbar, Brian. "Biomass Burning Fact Sheet." NASA, nasa.gov/centers/langley/news/factsheets/biomass.html.

"What Is ForestGEO?" ForestGEO, 15 Sept. 2020, forestgeo.si.edu/what-forestgeo.

"CZU Lightning Complex (Including Warnella Fire)." Cal Fire Department of Forestry and Fire Protection, fire.ca.gov/incidents/2020/8/16/czu-lightning-complex-including-warnella-fire/.

Richards, Sam. "CZU Lightning Fires to 74,000 acres in San Mateo, Santa Cruz counties." Piedmont Exedra, piedmontexedra.com/2020/08/czu-lightning-fires-to-74000-acres-in-san-mateo-santa-cruz-counties. Accessed 8 December, 2020.

Martin, Roy W. "RESOURCE INVENTORY: BIG BASIN REDWOODS STATE PARK AUGUST 1998." Associate Resource Ecologist Northern Service Center, parks.ca.gov/pages/21299/files/bbplant.pdf.

Hyndman, Rob J., and George Athanasopoulos. Forecasting: Principles and Practice. OTexts, 2018.

Dietze, Michael C. Ecological Forecasting. Princeton University Press, 2017.


Related Articles