Boreal forests are an important part of the global carbon cycle and recent climate change is endangering the large carbon stores in the soils and biomass of these ecosystems with dangerous consequences for future generations. A number of studies using time series data from the Advanced Very High Resolution Radiometer (AVHRR) have shown declining NDVI trends in the boreal forests of Canada that may be linked to changes in forest productivity. We have done research to investigate these trends using higher spatial resolution Landsat time series for the period 1984-2011. My role in this was to develop a procedure for processing a dense time series of Landsat data for the purpose of detecting trends in peak-summer greenness and the timing of phenological events. To do this, I chose a set of 46 Landsat scenes across the Canadian boreal forest sites that correspond to overlapping Landsat acquisitions. Using a maximum NDVI compositing procedure I created peak-summer greenness images for each year between 1984 and 2011. First, for fifteen of the sites, I demonstrated how data artifacts in the time series influence the detection of Landsat NDVI trends. Second, I examined how Landsat vegetation index (VI) trends are dependent on land cover, disturbance, and climate factors. These Landsat-derived trends will be useful to compare with those from AVHRR and MODIS time series that have poorer spatial and radiometric resolutions. Following my initial work we received a grant for the NASA ABoVE project to expand the work of detecting disturbance, identifying phenology dates, and detecting peak-greenness trends for all of Western Canada and Alaska.
Related Publications
Sulla-Menashe, D, MA Friedl, and CE Woodcock. 2016. Sources of bias and variability in long-term Landsat time series over Canadian boreal forests. Remote Sensing of Environment, 177: 206-219.
Project Summary
Landsat Analysis of Boreal Forest Change
External Links
USGS archive for downloading Landsat data
Landsat-based land cover map of Canada from the Earth Observation for Sustainable Development of Forests (EOSD)
Canadian National Fire Database from Natural Resources Canada website
A National Ecological Framework for Canada GIS layers
Canada's National Forest Inventory web portal
Tutorial for downloading and pre-processing Landsat stacks using Python/UNIX
Abstract:
A variety of evidence suggests that the boreal forests of Canada are responding to climate change. Specifically, several studies have inferred that widespread browning trends detected in time series of the Normalized Difference Vegetation Index (NDVI) from the Advanced Very High Resolution Radiometer (AVHRR) reflect the response of boreal forests to longer growing seasons, increased summer drought stress, and higher frequency of fires. Data from the Thematic Mapper (TM5) and Enhanced Thematic Mapper Plus (ETM+) sensors onboard Landsat 5 and 7, respectively, span essentially the same time period as the AVHRR record, but provide data with substantially higher radiometric and spatial fidelity, and by extension, a much improved basis for evaluating decadal-scale trends in spectral vegetation indices such as the NDVI. However, detection of trends, which are often subtle, requires careful attention to ensure that artifacts associated with the quality and stability of inter- and intra-sensor calibration do not lead to spurious conclusions in results from time series analyses. In this paper, we use time series of TM5 and ETM+ images for fifteen sites distributed across the Canadian boreal forest zone to explore if and how sensor geometry and inter- and intra-sensor calibration affect trends in spectral vegetation indices derived from multi-decadal Landsat time series. To do this, we created annual cloud-free composites for each Landsat spectral band based on peak summer NDVI at each site from 1984 to 2011 using all available TM5 and ETM+ data. To distinguish trends arising from long term climate change from those related to disturbance, we isolated areas within each site that were undisturbed during the Landsat record, and used these locations to analyze sources of variance in time series of red reflectance, near-infrared (NIR) reflectance, the NDVI, and the Enhanced Vegetation Index (EVI). Our results highlight the challenges involved in distinguishing trends in surface properties from data artifacts caused by undetected atmospheric effects, changes in sensor view angles, and subtle radiometric differences between the TM5 and ETM+ sensors. In particular, differences in sensor view geometry across adjacent overlapping Landsat scenes cause vegetated pixels in the eastern portion of Landsat scenes to have higher reflectances in the red and NIR bands (by 5 and 6 percent, respectively) than pixels in the western portion of scenes. While this effect does not significantly change NDVI values, it does affect EVI values. We also found modest, but potentially significant, differences between the red band reflectance of each sensor, with TM5 data having 14 percent higher red reflectance on average for vegetated pixels, which can introduce spurious trends in time series that combine TM5 and ETM+ data. More generally, the results from this work demonstrate that while the 30+ year Landsat archive provides unprecedented opportunities for studying changes to the Earth's terrestrial biosphere over the last three decades, care must be taken when inferring trends in these data without considering how sources of variance unrelated to surface processes affect the integrity of Landsat time series.
Sulla-Menashe, D, MA Friedl, and CE Woodcock. In review. Boreal Forest Greening and Browning is Primarily Caused by Disturbance, not Climate Change. Submitted to Environmental Research Letters.
Abstract:
Recent studies based on satellite-derived normalized difference vegetation index (NDVI) time series indicate that boreal forests have experienced widespread browning over the last three decades, and have suggested that these patterns reflect decreases in forest productivity induced by climate change. Using a large sample of NDVI time series with much higher quality and spatial resolution compared to previous studies, we show that the majority of NDVI changes in Canada's boreal forest are located in areas that experienced disturbance over the last three decades and reflect disturbance-recovery dynamics, not climate change impacts. NDVI data for undisturbed forests, where time series are not affected by the exogenous effects of fire, insects, and other forms of disturbance, exhibit local patterns of greening and browning that are consistent with regional changes in climate. However, these changes are modest, geographically limited, and depend on local bioclimatic conditions. Where present, NDVI changes unrelated to disturbance tended to be located near boundaries of the boreal forest bioclimatic envelope. We conclude that continued long-term climate change has the potential to significantly alter the character and function of Canada's boreal forest, but recent changes have been modest and near-term impacts are likely to be focused in or near ecotones.
The Canadian boreal forest study area and the 46 sites (identified by red boxes) selected according to (A) ecozones, (B) AVHRR NDVI trend slopes, (C) tree cover, and (D) fire disturbance.
Distribution of NDVI trends for the same sample of 4,400 undisturbed forest pixels. In the first distribution (combined), maximum-NDVI compositing was used to combine TM5 and ETM+ data for each year during the 1999-2011 period. In the second distribution (TM5-only), maximum NDVI compositing was used to create annual composites using only TM5 data between 1984-2011. Trends were detected using the Theil-Sen estimator and significance was assessed based on the Mann-Kendall test.
Comparisons of mean red reflectance, NIR reflectance, NDVI, and EVI values for undisturbed pixels between TM5 and ETM+ data. The values plotted are means of five year (2005-2009) medians, stratified by land cover and study site.
Boxplots for NDVI trends grouped by time since disturbance and forest type. The numbers to the right of the plot show the total area covered by Landsat pixels in each category.
The magnitude and direction of climate change at each study site as described by 28-year (1984-2011) changes to mean annual temperature and precipitation, along with the approximate location of climate envelopes for boreal forest and adjacent biomes according to Whittaker and adapted from (Gauthier et al., 2015). Each arrow shows the change in mean annual temperature and mean annual precipitation for that site, where the color indicates whether the site experienced greening or browning, and the width of the arrow indicates the magnitude of change in NDVI between 1984 and 2011.