Hosting Organization:  




news and stories about regional climate science


Urban Microclimates - Tracking Tree Seasonality in Boulder, CO

Urban Microclimates - Tracking Tree Seasonality in Boulder, CO

- Deidre Jaeger, NSF Graduate Research Internship Program fellow, in a PhD program at CU-Boulder

I’m interested in how tree phenology - the timing of seasonal activity such as budburst and leaf out - may be influenced by climate change and urban microclimates. Start of spring is particularly sensitive to temperature, one of the main drivers of spring budburst. The National Phenology Network (NPN) gathers observations of start of spring in the United States that have been mirroring the increase in average spring temperatures at the landscape scale.

Here's the problem: while there are 966 NPN observation sites in Colorado and 46 sites in Boulder, Colorado, very few are active. When thinking about the top three dominant trees in Boulder - Ash (Fraxinus sp.), Plains Cottonwood (Populus deltoides), and Siberian elm (Ulmus pumila) - there are only 12 sites where Eastern cottonwood has been monitored in Colorado and no observations of ash or Siberian elm. As a result, we don't have a good way to monitor changes in start-of=spring events here in Colorado. 

That's where my work comes in. I am observing tree phenology in Boulder at several scales. First, I make ground observations of ash trees, an introduced species in Colorado that is native to the Eastern US. Boulder is the first city to have Emerald Ash Borer detected in this state, and the City of Boulder urban forestry department is working to detect, treat, or remove infected trees. I am interested in the ecosystem services offered by these trees as they are the dominant species in Boulder’s public canopy. I am measuring the air temperature under the tree canopy with Hobo sensors at each of seven trees to start to get a sense of cooling benefits offered by shade and evaporation off the leaves.

I am also field testing the use of an accelerometer to give information about a single tree’s phenology. Last fall, I strapped accelerometers to seven ash tree trunks to measure the timing of end of season color change and leaf drop. The idea is centered on the relationship between acceleration and tree mass. As the tree mass changes over the growing season, the acceleration of the trunk vibrations in the wind will also change. I found this loss of mass trend in 6 of 7 of my trees, with different degrees of signal strength, as the trees changes color and lost their leaves. Brian Miller of the NCCSC and Jonathan Friedman of the USGS Boulder/Ft Collins, have been helping me design this project and interpret results. I am super excited to keep developing a way to measure tree phenology of individual trees at a larger scale! 

To that end, I'm also using remote sensing in my work. Remote sensing from satellites offers powerful spatial and temporal coverage of land surface phenology and microclimates. I am interested in what satellite-derived data from vegetation indices (such as NDVI) and land surface temperature can tell us about urban heat islands and urban phenology shifts. Remote sensing has been useful for detecting urban heat island effects, which can occur where more radiation is absorbed by city buildings and pavement compared to surrounding rural areas. The urban heat island can lead to extreme temperatures, posing heat stress risk to human and animal health. Denver has a strong urban heat island effect, but less is known about this effect in other smaller Front Range Cities that are still undergoing urbanization and increased land development pressure. Dr. Gabriel Senay of the NCCSC is helping me use remote sensing data to look at whether Boulder has an urban heat island effect and how vegetation phenology has changed over time with increased land-use. I am comparing NDVI, a measure of light reflected by photosynthetically active plants, collected from the Modis satellite to my ground measurements. Even though the Modis satellite passes over the same spot every 1-2 days, clouds can obstruct the imagery and so it can be hard to get enough clear images to look at changes within a short period of time. For this project, remote sensing NDVI and land surface temperature composites (best images from 16 days) may be most useful for getting a sense of general temporal or spatial shifts happening at the landscape level. 

: An accelerometer, which can be thought of as a "fitbit for trees," tracks tree sway and its relationship to tree mass, which changes during the growing season. This relationship allows a researcher to remotely track when trees leaf out and when they lose their leaves. 

Deidre Jaeger is working with Jonathan Friedman (USGS), Brian Miller (USGS, NC CSC) and Gabriel Senay (EROS, NC CSC) as mentors for this program.