Doctor of Philosophy (PhD)


Renewable Natural Resources

Document Type



This dissertation research focused on three questions: (1) what is the current carbon stock in Louisiana’s forest ecosystems? (2) how will the biomass carbon stock respond to future climate change? and (3) how vulnerable are the coastal forest resources to natural disturbances, such as hurricanes? The research utilized a geographic information system, remote sensing techniques, ecosystem modeling, and statistical approaches with existing data and in-situ measurements. Future climate changes were adapted from predictions by the Community Climate System Model on the basis of low (B1), moderate (A1B), and high (A2) greenhouse gas emission scenarios. The study on forest carbon assessment found that Louisiana’s forests currently store 219.2 Tg of biomass carbon, 90% of which is stored in wetland and evergreen forests. Spatial variation of the carbon storage was mainly affected by forest biomass distribution. No correlation was identified between carbon storage in watersheds with the average watershed slope and drainage density. The modeling study on growth response to future climate found that forest net primary productivity (NPP) would decline from 2000 to 2050 under scenario B1, but may increase under scenarios A1B and A2 due primarily to minimum temperature and precipitation changes. Uncertainties of the NPP prediction were apparent, owing to spatial resolution of the climate variables. The remote sensing study on hurricane disturbance to coastal forests found that increases in the intensity of severe weather in the future would likely increase the turn-over rate of coastal forest carbon stock. Forest attributes and site conditions had a variety of effects on the vulnerability of forests to hurricane disturbance and thereby, spatial patterns of disturbed landscape. Soil groups and stand factors, including forest types, forest coverage, and stand density contributed to 85% of accuracy in the modeling probability of Hurricane Katrina disturbance to forests. In conclusion, this research demonstrated that quantification of forest biomass carbon, using geo-referenced datasets and GIS techniques, provides a credible approach to increase accuracy and constrain the uncertainty of large-scale carbon assessment. A combination of ecosystem modeling and GIS/Remote Sensing techniques can provide insight into future climate change effects on forest carbon change at the landscape scale.



Document Availability at the Time of Submission

Release the entire work immediately for access worldwide.

Committee Chair

Y. Jun Xu