research interests

Research Interests

Ecology:

forest structure and dynamics, fire ecology, biodiversity, biogeography, macroecology

Remote sensing:

lidar, image spectroscopy, broad-band optical time series

GIScience:

geospatial analysis, ecoinformatics, data visualization

Statistics:

multivariate, spatial, hierarchical, nonparametric, prediction and inference

Applications:

conservation, wildfires, human-environment, land cover change, urbanization

Screen Shot 2023-01-05 at 11.58.53 AM.png

Earth Dynamics Geodetic Explorer (EDGE)

EDGE seeks to understand how the 3D structure of terrestrial ecosystems and the surface topography of glaciers, ice sheets and sea ice are changing in response to climate and human activity. In 2023, EDGE was down-selected among four proposed NASA Earth System Explorers (ESE) Program – which conducts principal investigator-led space science missions as recommended by the National Academies of Sciences, Engineering, and Medicine 2017 Decadal Survey for Earth Science and Applications from Space – for concept studies of missions to help us better understand Earth science key focus areas for the benefit of all including greenhouse gases, the ozone layer, ocean surface currents, and changes in ice and glaciers around the world. The project, which includes GEODE Assistant Research Professor Chris Hakkenberg as Co-I and Biodiversity Co-Lead, will receive $5 million to conduct a one-year concept study. Should the team’s concept be one of two of the four teams selected to advance to the final stage, they will receive a $310 million budget to build the instrument and, with the help of NASA, launch it into space in 2030.
With an increased density of laser beams that will map the planet in five 120-meter-wide strips, EDGE builds on and advances space-borne lidar technology like GEDI and ICESat-2. In comparison, EDGE will provide much higher resolution and accuracy – with vertical accuracy < 1m and contiguous horizontal sampling in 120m swaths with <3m geolocational accuracy – and a much larger extent, ±83 deg latitude. This will provide a more comprehensive view of the dynamics of the Earth’s surface. For Terrestrial Ecosystem Structure, EDGE will (1) quantify how the 3D vegetation structure and aboveground biomass of Earth’s wooded ecosystems has changed in response to natural and anthropogenic disturbances; (2) quantify how 3D vegetation structure and surface topography are related to habitat suitability and biodiversity for the Earth’s wooded ecosystems; (3) quantify woody vegetation carbon fluxes and sequestration potential in response to land use and climate change.

Quantifying carbon associated with Redwoods Rising

This three-part study seeks to assess carbon (C) storage and annual C sequestration in Redwood National and State Parks (RNSP), specifically via the Save the Redwood League’s Redwoods Rising restoration activities in Humboldt and Del Norte Counties, CA. Redwoods Rising seeks to implement mechanical thinning in second-growth redwood forests to accelerate the development of old-growth structures. This study will utilize local expertise from the USGS Western Ecological Research Center and Cal Poly Humboldt (CPH) in conjunction with remote sensing mapping and C modeling capacities from NAU. The project combines (1) tree, plot, and stand allometric and biomass relationships; (2) stand development models; (3) air- and space-borne remotely-sensed data; and (4) advanced machine-learning modeling procedures to generate spatially-explicit annual maps of forest C storage and annual sequestration across a range of stand conditions including primary (old-growth) forests, unmanaged secondary (second-growth) forests, and managed secondary forests in the RNSP study area. Outputs from the study will be used to assess the feasibility of a carbon-trading scheme to eventually sustainably fund Redwoods Rising’s operations.

Using GEDI to assess fuel structure, fuel treatments, wildfire severity and post-fire transitions across California wildfires

We are working with with academic, state, and federal partners to employ hindcasted and interpolated space-borne lidar, with field and airborne data to assess the relationship between pre-fire fuels, weather and climate conditions, fire severity, and fuels treatment effectiveness across America’s West, and especially in the wildland-urban interface (WUI). This multi-project study consists of parts. In the first part, funded from a CALFIRE Forest Health Research grant, investigators will (1) assess whether pre-fire GEDI structural metrics increase the accuracy of forest wildfire severity predictions relative to other commonly-used predictor variables; (2) use GEDI to assess structural change from wildfires as well as from mechanical treatments; (3) interpolate GEDI to create continuous wildfire risk maps across California; and (4) compare predicted wildfire severity among CALFIRE’s Reduce Wildfire Threat to Communities (RWTC) and Reducing Wildfire Risk to Forest Ecosystem Services (RWRFES) priority landscapes. In the second stage, funded with a NASA Global Ecosystem Dynamics Investigation (GEDI) Science Team grant, investigators will derive and validate horizontal landscape maps from GEDI vertical profiles, compare GEDI-derived ‘quantity’ versus optical-derived ‘quality’ of canopy fuels in driving wildfire severity; (3) use GEDI to improve optical-based burn severity maps; and (4) employ multi-temporal continuous GEDI maps of structure to characterize decadal transitions in vegetation structure and pattern for >100ha wildfires across three ecoregions of California (North Coast, Central Coast, and Sierras) between 2006-2009 and 2019-2022.

Global forest community dynamics

My work on global-scale forest community dynamics relies on a diverse array of data spanning spatial and temporal scales, including high spatial resolution optical imagery, image spectroscopy (hyperspectral), LiDAR, and ground plots to model interrelations among emergent properties of forest ecosystems like biodiversity, composition, structure, and function. Space-borne sensors like GEDI, DESIS, ECOSTRESS, and upcoming SBG coupled with global networks of airborne LiDAR-hyperspectral networks like NEON and G-LiHT make scaling properties from landscapes to global scales increasing feasible.

Climate, environment, structure, and diversity

I am employing satellite and aerial data to examine environmental and climatological constraints on the relationship between vegetation structure, diversity, and ecosystem function across the biomes of North America. Results will assist collaborative efforts to model multi-scale community processes at a global extent using space-borne LiDAR and hyperspectral sensors like GEDI and SBG.

Screen Shot 2020-02-22 at 3.36.55 PM.png

Ecological scaling

Ecological phenomena are fundamentally scale dependent in that they vary as a function of biological, spatial, and temporal grain and extent. Inferential conclusions from any one scale of observation are potentially misleading when applied to another. For this and other reasons, multiple-scale analyses are better suited to uncovering the combined effect of multiple drivers of diversity acting at multiple scales. One fundamental pillar of my research approach is to systemically assess the role of scale: including the characteristic scales at which organisms disperse, assemble, and compete, considerations of resolution and extent in multiple sources of remotely-sensed data, to tradeoffs in parsimony and complexity in models of ecosystem properties.

Screen Shot 2020-02-22 at 2.54.47 PM.png

Other research

My work on land cover change time series employs repeat imagery from constellations of space-borne sensors to create large-extent, temporally-dense multi-decadal land cover time series. My interest in the topic is both methods-driven and applied. I have developed a series of algorithms for automating data fusion and predictive modeling of multi-temporal imagery data cubes. In addition, I led studies to characterize higher-order spatio-temporal land cover change morphologies in the Greater Houston area (summarized here).

NASA Earth and Space Science Fellowship. In this three-part study, I use field plots, high spatial resolution optical imagery, image spectroscopy, and laser altimetry from NASA Goddard’s LiDAR, Thermal, and Hyperspectral (G-LiHT) airborne imager to map forest communities over a continuous environmental gradient and predict vascular plant species richness at seven spatial scales using predictive Bayesian spatial modeling.

NSF East Asia and Pacific Summer Institutes grant. After pervasive logging and other forms of environmental degradation over the past thirty years, NFPP forest coverage is increasing in the landscape of NW Yunnan, China. As a space-for-time substitution, results from this study indicate a temporal pattern of enhanced recruitment of historically present but currently rare tree species in recently converted secondary forests in proximity to refugial, late successional Tibetan sacred forests. This process is mediated by changing biotic and abiotic conditions along a spatio-temporal gradient. Conservation born of religio-cultural tradition as observed in Tibetan village sacred forests provide a rich example of the conservation outcomes of centuries-old socio-ecological practices and the role of sacred forests as island refugia harboring rare plant genotypes amid large-scale ecological degradation.

Spatially Explicit Plot-Level Forest Dynamics: An Interactive Tool. Fast and flexible three-dimensional forest dynamics spatio-temporal data visualization. Provides ecological statistics based on selected criteria, spatially-explicit forest dynamics map, and change vectors. Requires one .txt file with three data points per tree: species type, diameter at breast height (dbh), xy location.

Luce Foundation Grant for Chinese Studies. The Lost Voices: The Impact of the Development and Preservation Policies upon the Local People of Wudang Shan. This research focuses on over forty interviews concerning contested development and environmental policies in the UNESCO World Heritage Site: Wudang Shan in Hubei, China.