Research Details

Climate Extremes

The NSF Macrosystems Biology project is a joint effort between Lehigh, MIT, MBL, and UC-Davis. The purpose of our project is to explore the role of climate extremes (floods, droughts) on ecosystem functions and services, with a particular focus on non-linear threshold effects. This program is partly in response to the new system of eddy covariance flux towers within the National Ecological Observatory Network (NEON), so we are using data from the existing Ameriflux network as our primary data source. I have developed a daily version of TEM-Hydro, which is forced with daily climate data rather than monthly, in order to better model the effects of both floods and droughts on ecosystem and vegetation productivity and other ecosystem services like runoff and crop yield. A recent publication in a special MSB issue of Frontiers in Ecology and the Environment (Ruegg et al., 2013) discusses the issue of data informatics. Both my Ph.D. students are studying the effects of climate extremes on ecosystems, with Mingkai focusing on ecosystem tipping points resulting from climate extremes in the Columbia River Basin and Jien focusing on the effect of drought on recharge of the Ogellala Aquifer in the Great Plains. Ph.D. student Nicolas Bambach at UC Davis is currently working to couple the Advanced Canopy-Atmosphere-Soil Algorithm (ACASA) model with the TEM-Hydro model, to incorporate carbon and nitrogen dynamics into the existing plant-canopy model.

Figure 1
Lehigh University Benjamin Felzer - Vegetation Carbon: Pine

Disturbance scenarios effect on vegetation carbon at 5 temperate coniferous sites, as described in table.

Figure 2
Lehigh University Benjamin Felzer - NEP: Duke Forest and NEP NC Loblolly Pine

Effect of land use change and disturbance on carbon fluxes at Duke and NC, both loblolly pine sites. Duke was affected by timber harvest in 1982 and an ice storm in 2002; NCL was abandoned agriculture in 1968 and cut by timber harvest in 1992. A younger regrowing forest is a stronger carbon sink than a more mature forest, so it is important to accurately capture the effects of disturbance. During timber harvest, material removed from the site will not release carbon at the site and so will strengthen the carbon sink. During storms, a significant portion (2/3) of the stem material decomposes directly to the atmosphere in the years immediately following disturbance, while the remainder enters a standing dead pool where it slowly enters the soil through slash over time. A forest site on former agriculture land will have more nutrients due to fertilization during cropping. As these disturbed systems regrow, they sequester more carbon due to more rapid growth rate.

Figure 3
Lehigh University Benjamin Felzer - ET Blodget Forest and NEP Blodget Forest

The type of disturbance is also important. In these experiments at Blodget Forest, the timber harvest in 1990 has been replaced by a storm, which has higher ET and NEP in the earlier years closer to the disturbance event, but fluxes are similar with more time.  Besides removing carbon from the soil, the loss of nitrogen during timber harvest makes the system more nitrogen-limited, which reduces the carbon sink. Timber: create wood products and clearcut (stand-replacing); Storm: 8% of roots + labile and 41% of leaves in slash; 8% of stems in standing dead.

Figure 4
Lehigh University Benjamin Felzer - ET Harvard Forest and NEP Harvard Forest

Run with and without nitrogen deposition for Harvard Forest showing effect of N-limitation on forest growth.

Flood Study

Because local land use planners are currently lacking the kind of climate and ecosystem scenarios at the regional scale that will allow them to plan for or adapt to the impact of climate variability and extremes in the future, I have been exploring how to scale between low resolution climate models and high resolution hydrological models used by water managers.  I supervised two undergraduate students, Lauren Schneck and Cathy Withers, in using 20th and 21st century hourly precipitation values generated by our CESM runs to determine the increase in storm intensity in the Lehigh Valley, and then applied the storm statistics to the U.S. Army Corp of Engineers - Hydrologic Engineering Center – Hydrologic Modeling System and River Analysis System (HEC-HMS and HEC-RA) for part of the Monocacy basin.  These models are used by watershed managers and the Federal Emergency Management Agency (FEMA) to determine peak discharge, water surface levels, flood inundation, and flood insurance rate maps.  Such an effort can form the basis for allowing decision-makers to integrate modeling research into regional planning documents and decisions.  We have recently submitted an article to Journal of Flood Risk Management (Felzer and Withers) to describe the results of this research.

Figure 9
Lehigh University Benjamin Felzer - Historic Return Intervals and Return Intervals

Converting rainfall to discharge, the historical 100-year event was is 2920 ft3s-1but will become 3114 ft3s-1 in the future. 

Felzer, B.S. and Withers, C.E. Using future storm statistics from climate models to determine flood potential in the Lehigh Valley, PA in the 21st century. Journal of Flood Risk Management.

Effect of climate variability on human civilization: I was just awarded an NSF IBSS proposal with a team of anthropologists to explore the human and cultural response to climate disasters.  The key issue is how climate disasters such as droughts and floods affect food supply for a global set of preindustrial, prehistoric, and contemporary societies.  This research will help us understand how human societies have responded and adapted to differing types, frequencies, and predictabilities of climate-related extremes.