Woods Hole Research Center
Woods Hole, MA 02543-0296
Program Manager: Kathy Tedesco NOAA/OGP Global Carbon Cycle Program
Related NOAA Strategic Plan Goal:
Goal 2. Understand climate variability and change to enhance society’s ability to plan and respond.
This project was focused on the analysis of North American boreal forest carbon dynamics. There were several aspects to the project, including: (i) time series analysis of NOAA Advanced Very High Resolution Radiometer (AVHRR) imagery on 10-15 day intervals over the past two decades, (ii) estimation of biophysical attributes of the vegetation, including canopy light harvesting as it relates to photosynthesis over large areas, (iii) use of the image data, along with ancillary data sets, to track seasonal and interannual variability in burned, unburned and regenerating areas, (iv) collection and analysis of field data sets to calibrate and validate the results. The data sets that we produced and the analyses that we conducted have direct relevance to the research objectives of NOAA’s Climate and Global Change Program, and the USGCRP North American Carbon Program.
The results of much of this work summarized here are also provided on the internet at www.whrc.org/borealnamerica. Our estimates of vegetation processes from satellite observation were based on field measurement activities in interior Alaska. We describe the indirect estimation of leaf area index (LAI) from handheld instruments at 34 sites in the study area, including factors which influence the uncertainty of estimates (Hyer and Goetz 2004, listed under publications). Our related continuous field measurements of canopy light interception (Fpar) were processed for the period 2001 through 2004, with careful sensor cross-calibration. We used these data to analyze temporal variation in light harvesting, for validation of estimates from satellite observations (Figure 1), and for scaling the observations to other satellite data sets of different spatial resolution. We published on this component of the project in a special issue on MODIS validation (Steinberg et al. 2006, see publications).
The time series analysis of AVHRR satellite data sets (64 km2) was conducted across boreal North America following processing of the data sets (described in the Year 1 report). The normalized difference reflectance index (NDVI) data set was converted to estimates of the fraction of incident photosynthetically active radiation (PAR) intercepted by vegetation (i.e., Fpar), and these data were then analyzed across a wide range of burned areas across boreal North America.
Regeneration from fires that burned in three episodic fire years (1981, 1989, and 1995) was characterized, including the gradual increase in canopy light harvesting (Fpar) as vegetation density and stature increased through time after fire (Figure 2). The rate of recovery varied in different fire years, even for the three episodic fire years on which we focused our analysis. Variability in the time series remaining after accounting for environmental factors, represented by the difference of burned and unburned area anomalies, increased beyond the observed postfire recovery period which may indicate residual effects of fire disturbance over the regrowth period. As part of this analysis we also conducted a comparison of different AVHRR data sets. The results of this component of the project were also published (Goetz et al. 2006, see publications).
The same AVHRR data sets were used to conduct a rigorous time series and trend analysis of photosynthetic activity (Pg) over the 22 year period across boreal North America, where Pg was estimated from Fpar (above). The trend analysis, which focused on unburned areas in order to characterize ecosystems responses other than disturbance, revealed that tundra vegetation experienced an increase in both peak photosynthesis and growing season length, whereas many forested areas actually experienced a decline in photosynthetic activity between 1981 and 2003 (Figure 3). Climatic warming occurred across the entire region (Figure 4), but the change in the
forest response indicates that long-term changes may not be predictable from initial, short-term observations. Fire disturbance also increased with the warming (see inset of Figure 3) but did not explain the decline in forest photosynthetic activity.
This work represents a potentially important advance at broad spatial scales by making use of the full extent of the satellite observational record to document unique vegetation responses to climatic warming. The results are striking, unexpected, and at odds with some earlier work that indicated ubiquitous greening. Our results were published in the Proceedings of the National Academy of Sciences (Goetz et al. 2005, see publications), and were picked up by a number of public media outlets. NASA’s “Earth Observatory” and other organization published features on the work (summarized below). These results provided some of the first evidence that high
latitude forests may be in decline following an initial greening trend associated with CO2 and warming. These trends have important implications for the direction of feedbacks to the climate system through changes in surface albedo and, when considered with our related work, fire disturbance regimes. This is particularly important since climate models predict continued warming and interior drying across the region.
We considered a number of possible causes for these observations, and extended the analysis in three ways: (i) by analyzing results specific to cover type, canopy density, and timing (seasonality) of the observed trend, (ii) by analyzing the primary climate variables associated with observed photosynthetic responses, including temporal lags in responses, (iii) by analyzing the AVHRR data sets across the entire circumpolar high latitude region. The results of these analyses, which were summarized in Bunn et al. (2005) and Bunn and Goetz (2006) (see publications), improved our understanding of the processes involved. Specifically, we found that (i) trends were strongest in late summer, indicating that drought stress (warmer and dryer air masses) was an important factor, (ii) areas of greater tree canopy density were more negatively impacted than sparse tree cover and/or more herbaceous areas, (iii) vegetation cover types keyed on different climate variables and had associated differences in their lagged responses to climate (figure 5), (iv) the observed trends held for high latitude forests of northern Eurasia as well as North America.
We have proposed to examine these factors in greater detail as part of follow-on work (proposal pending under FY07 NOAA OGP announcement of opportunity).
Publications Resulting from this Project:
Bunn A. G., Goetz S. J. & Fiske G. J. (2005). Observed and predicted responses of plant growth to climate across Canada. Geophysical Research Letters. 32:doi:10.1029/2005GL023646
Bunn, A. G., and S. J. Goetz. (2006) Trends in circumpolar satellite observed gross photosynthesis from 1982-2003: The role of cover type and vegetation density. Earth Interactions 10(12):1-19.
Goetz S. J. Fiske G. & Bunn A. (2006). Using satellite time series data sets to analyze fire disturbance and recovery in the Canadian boreal forest. Remote Sensing of Environment 101 (3):352-365.
Goetz S. J., Steinberg D., Fiske G. & Houghton R. A. (2005) Satellite observed photosynthetic trends across boreal North America associated with climate and fire disturbance. Proceedings of the National Academy of Sciences 102(38): 13521-13525.
Hyer, E., and S. J. Goetz (2004). Comparison and sensitivity analysis of instruments and radiometric methods for LAI estimation: assessments from a boreal forest site. Agricultural and Forest Meteorology 122 (3/4):157-174.
Kasischke, E. S., S. J. Goetz, M. Hansen, M. Ozdogan, J. Rogan, S. L. Ustin, and C. E. Woodcock. (2004). Temperate and Boreal Forests. Pages 147-238 in S. L. Ustin, editor. Remote Sensing for Natural Resource Management and Environmental Monitoring. co-published by John Wiley & Sons and American Society of Photogrammetry and Remote Sensing, Hoboken, NJ.
Steinberg D. C., Goetz S. J. & Hyer E. (2006) Validation of MODIS Fpar products in boreal forests of Alaska. IEEE Transactions on Geoscience and Remote Sensing 144(7):1818-1828.
Ecosystem responses to high latitude warming, Public Symposium at Cape Cod Community College (March 2006), Barnstable MA
Trends in Circumpolar Photosynthetic Activity from 1982-2003, Arctic Research Consortium of the United States (May 2006) and American Institute of Biological Sciences (May 2006), Washington DC
Photosynthetic trends in the high latitudes: Is the boreal forest greening or browning? AGU Fall Conference (Dec 2005), San Francisco
Mapping and Monitoring Boreal Forest Regrowth Dynamics using Satellite Data Products, AGU Fall Conference (Dec 2004), San Francisco.
Remote Sensing and Modeling of Boreal Forest Regrowth Dynamics, International Boreal Forest Research Association (IBFRA) Conference, Climate-Disturbance Interactions in Boreal Forest Ecosystems (May 2004), Fairbanks, Alaska.
Other venues (selected)
American Museum of Natural History, Browning boreal forests, http://sciencebulletins.amnh.org/bio/s/boreal.20060508/index.php
Anchorage Daily News (front page feature), 15 Sept 2005, Arctic tundra growing greener as northern forests dry out, http://www.adn.com/front/story/6970702p-6870818c.html
Dozens of other periodicals picked up portions of releases by the Associated Press and United Press International, including Climate change in high latitudes studied,