Research



Trained as a biometeorologist, my research is centered within the area of land-atmosphere

interactions. The overall question that has motivated my research is: what is the impact of

land surface heterogeneity onboundary layer processes, in particular the cycling of

water, carbon and energy? Fundamentally, this is a question of scaling, and the answer is

essential for increasing our understanding and ability to model the Earth system and ultimately

for mitigating the impacts of future climate change. Specic examples of my recent studies

include the impact of the spatial scale of land surface heterogeneity on modeled uxes of water

and carbon, examining the role of changing regional climate on ecophysiological processes in

the central U.S. and assessing the ability of satellite data to assess urban heat islands and

validate regional and global climate model simulations of urban heat stress. I approach these

studies using a combination of techniques including remote sensing analysis, eld observations

and numerical modeling.

Prior to tenure, I examined the impacts of climate change in the central U. S. (Brunsell

et al. 2010). This work was motivated by an interest in understanding land cover variability

in response to global climate change. This work compared GCM output with 20th century

meteorological observations to quantify the ability of GCMs to capture 20th century conditions

in the central U.S. Since tenure, my focus has shifted to assessing the phenological implications

of this change in climate. Working with a low dimensional modeling framework, a graduate

student and I examined the impact of changing precipitation on the ecohydrology in the central

U.S. (Petrie and Brunsell, 2012) as well as extending this to examine the implications of

changing climate on coupled carbon and water dynamics (Petrie et al., 2012). This led me to

the issue of extreme events and questioning the interaction between the timing of the event

versus the magnitude of the event (Craine et al., 2012). We are currently developing an analysis

which we call the Critical Climate Period (CCP) approach that highlights the sensitivity to

meteorological anomalies as a function of time of year and length of the event. I have begun to

examine to what extent ecological process models (e.g. Biome-BGC) and land surface models

(e.g. NOAH) are capable of capturing the sensitivity to the timing of extreme events. I envision

this line of research to be a major focus of mine in the next few years.

During my sabbatical year spent as a Humboldt Fellow at the Max Planck Institute for

Biogeochemistry in Germany, I further developed the technique of applying multi-scale informa-

tion theory methods to quantify the role of land surface heterogeneity on the land-atmosphere

exchanges of mass and energy (Brunsell and Anderson, 2011; Brunsell et al. 2011). This work

has been focused on quantifying how the information content in remotely sensed land surface

temperature and vegetation indices contributes to the information content in modeled surface

energy balance uxes. Recently, I have extended this methodology to examine the dierences

in information content between observed and modeled uxes as a function of time scale and

the contribution of the environmental variables such as soil moisture and air temperature to

the information content of the uxes (Brunsell and Wilson, 2013). My current work implies

that the observations and models diverge in their information contents at the multi-week scale,

thus implicating soil moisture dynamics as a major factor in model-data disagreement. Future

work in this area will focus on using the information theory metrics as a convenient method-

ology for diagnosing the temporal and spatial scales at which models have limited agreement

with observations. This will allow us to focus our model improvement eorts on the processes

dominant at those scales.

Another major focus of my recent research has been a developing collaboration with NCAR 

scientists to examine urban heat islands (UHI) and exposure to heat stress. This work has

been funded through NASA to examine the spatial and temporal dynamics of the UHI in the

cities of Houston and Toronto. A PhD student working under my direction, Leiqiu Hu, has

been successful in illustrating the bias induced by using temporally aggregated land surface

temperature for UHI statistics (Hu and Brunsell, 2013). In addition, we have a series of

papers highlighting the role of remotely sensed land surface temperature data for validating

and comparing regional (Hu et al., 2014; Monaghan et al., 2014) as well as global climate model

simulations of UHI dynamics (Oleson et al., in press). This work has now been extended to

examining the impacts of increased urbanization and climate change through a grant from

NSF (EASM2, PI: Brian O'Neill). Specically, we are examining urbanization and climate

change in Brazil, India, and China through an integrated approach that combines economic

simulations, population growth modeling, global climate models and remotely sensed data for

validation.

One question that has always interested me is: what is the role of the local microclimate

versus the role of regional climate on mass and energy cycling? This question has been one

of the primary reasons that I began maintaining eddy covariance towers on dierent land

cover types in Kansas. Currently, I actively maintain 4 eddy covariance stations as part of an

Ameriux Core Site as well as a large aperture scintillometer. These data have been essential

in assessing the role of land cover and regional climate change on land surface uxes (e.g.

Brunsell et al. 2014). The ux tower data has allowed me to become actively involved in

a number of studies such as examining the role of drought on grasslands (Shi et al., 2014),

and providing the validation data for satellite products (Campbell et al., 2013), and validating

model and remotely sensed estimates of surface uxes (Liu et al., in review). To extend the

remote sensing analysis to the question of regional versus local inuence, we have utilized the

tower data to examine the agreement between tropospheric CO2 concentration and local and

satellite observations (Cochran and Brunsell, 2012; Cochran et al., 2013). The local versus

regional impacts have also been investigated using the WRF regional climate model to assess

the impact of central plains irrigation (Huber et al. 2014) as well as interactions between

agricultural lands and precipitation interactions (Lin and Brunsell, 2014). Extending to a

larger spatial extent, we recently received word that a NASA/USDA proposal synthesizing

ux tower and satellite data to assess model uncertainty across the U.S. and Mexico has been

funded (PI: R. Vargas).


© Nathaniel Brunsell 2015