Crystal L. Schaaf

Global land cover mapping from MODIS: algorithms and early results

Until recently, advanced very high-resolution radiometer (AVHRR) observations were the only viable source of data for global land cover mapping. While many useful insights have been gained from analyses based on AVHRR data, the availability of moderate resolution imaging spectroradiometer (MODIS) data with greatly improved spectral, spatial, geometric, and radiometric attributes provides significant new opportunities and challenges for remote sensing-based land cover mapping research. In this paper, we describe the algorithms and databases being used to produce the MODIS global land cover product. This product provides maps of global land cover at 1-km spatial resolution using several classification systems, principally that of the IGBP. To generate these maps, a supervised classification methodology is used that exploits a global database of training sites interpreted from high-resolution imagery in association with ancillary data. In addition to the IGBP class at each pixel, the MODIS land cover product provides several other parameters including estimates for the classification confidence associated with the IGBP label, a prediction for the most likely alternative class, and class labels for several other classification schemes that are used by the global modeling community. Initial results based on 5 months of MODIS data are encouraging. At global scales, the distribution of vegetation and land cover types is qualitatively realistic. At regional scales, comparisons among heritage AVHRR products, Landsat TM data, and results from MODIS show that the algorithm is performing well. As a longer time series of data is added to the processing stream and the representation of global land cover in the site database is refined, the quality of the MODIS land cover product will improve accordingly.

Finding Leaves in the Forest: The Dual-Wavelength Echidna Lidar

The dual-wavelength Echidna lidar is a portable ground-based full-waveform terrestrial scanning lidar for characterization of fine-scale forest structure and biomass content. While scanning, the instrument records the full time series of returns at a half-nanosecond rate from two coaligned 5-ns pulsed lasers at 1064 and 1548 nm wavelengths. Leaves absorb more strongly at 1548 nm compared to stems, allowing discrimination of forest composition at milliradian scales from the ground to the forest canopy. This work describes the instrument design and data products and demonstrates the power of two wavelength lidar to clearly distinguish leaves from woody material with preliminary field data from the Sierra Nevada National Forest.

Estimation of the parameter error propagation in inversion based BRDF observations at single sun position

In multiangular remote sensing observations, the variable range of the solar zenith angle (SZN) is narrow. We takes the linear kernel-driven model as an example to analyze the parameter error propagation in inversion by using the observations at single sun position and to show that in such case it is unreliable to invert the BRDF tendency with SZN. To improve the algorithm, we suggest adding certain constraint for the changing trend of albedo with SZN in the model inversion.

Increased water yield due to the hemlock woolly adelgid infestation in New England

Over the past few decades, a hemlock woolly adelgid (HWA) infestation has significantly affected eastern hemlock (Tsuga canadensis) in the eastern U.S., and warmer winters are expected to promote a continued northward expansion in the future. Here we report a water yield increase due to the HWA infestation in New England, U.S. Since the first observation in 2002, peak growing season evapotranspiration over a hemlock-dominated area has decreased by 24–37% in 2012 and 2013. Over the same time period, the water yield from the study catchment significantly increased as compared to an adjacent catchment with less hemlock cover. The net increase was estimated to be as much as 15.6% of annual water yield in 2014 based on an ecohydrological modeling analysis. This study indicates that the ongoing hemlock decline is also largely altering hydrological regimes in the northeastern U.S.

Warming-Induced Earlier Greenup Leads to Reduced Stream Discharge in a Temperate Mixed Forest Catchment

The phenological response of vegetation to ongoing climate change may have great implications for hydrological regimes in the eastern United States. However, there have been few studies that analyze its resultant effect on catchment discharge dynamics, separating from dominant climatic controls. In this study, we examined the net effect of phenological variations on the long-term and interannual gross primary production (GPP) and evapotranspiration (ET) fluxes in a temperate deciduous forest, as well as on the catchment discharge behavior in a mixed deciduous-conifer forest catchment. First, we calibrated the spring and autumn leaf phenology models for the Harvard Forest in the northeastern United States, where the onsets of greenup and senescence have been significantly advanced and delayed, 10.3 and 6.0 days respectively, over the past two decades (1992–2011). We then integrated the phenology models into a mechanistic watershed ecohydrological model (RHESSys), which improved the interannual and long-term simulations of both the plot-scale daily GPP and ET fluxes and the catchment discharge dynamics. We found that the phenological changes amplified the long-term increases in GPP and ET driven by the climatic controls. In particular, the earlier greenup onsets resulted in increases in annual ET significantly, while the delayed senescence onsets had less influence. Consequently, the earlier greenup onsets reduced stream discharge not only during the growing season but also during the following dormant season due to soil water depletion. This study highlights the importance of understanding vegetation response to ongoing climate change in order to predict the future hydrological nonstationarity in this region.