Aaron Moody

Land-Surface Phenologies from AVHRR Using the Discrete Fourier Transform

The first and second harmonics of the discrete Fourier transform (DFT) concisely summarize the amplitude and phase of annual and biannual signals embedded in time-series of Advanced Very High Resolution Radiometer (AVHRR) Normalized Difference Vegetation Index (NDVI) data. We applied and evaluated the DFT using monthly composited NDVI data over a 7-year period for a 150 × 150-km study area in southern California. The study area contains strong gradients in environmental conditions and basic vegetation formations. Analysis of the DFT harmonics for six point-locations provided a basis for linking the DFT results to basic vegetation types according their characteristic phenologies. The mean NDVI, or 0th-order harmonic, indicated overall productivity, allowing the differentiation of unproductive, moderately productive, and highly productive sites. The amplitude of the first harmonic indicated the variability of productivity over the year as expressed in a single annual pulse of net primary production. This summarized the relative dominance of evergreen vs. deciduous or annual habit. The phase of the first harmonic summarized the timing of green-up relative to the timing of winter and spring rains. This differentiated rapidly responding annual grasslands, slowly responding evergreen life-forms, and irrigated agriculture. The second harmonic indicated the strength and timing of any biannual signal. This provided information on secondary vegetation types, such as subcanopy grasses beneath evergreen woodlands or mixtures of annual grasslands and irrigated agriculture. The point-based analysis provided the foundation for a regional analysis of the entire study area. The mean NDVI and first- and second-order amplitude and phase, in conjunction with 148 field-sampled polygons, were used to produce an unsupervised classification of basic vegetation formations for the study area. These results were evaluated by comparison with other land cover products, and through assessment using field-sampled test regions.

The influence of scale and the spatial characteristics of landscapes on land-cover mapping using remote sensing

Statistical analyses provide a means for assessing relationships between landscape spatial pattern and errors in the estimates of cover-type proportions as land-cover data are aggregated to coarser scales. Results from a multiple-linear regression model suggest that as patch sizes, variance/mean ratio, and initial proportions of cover types increase, the proportion error moves in a positive direction and is governed by the interaction of the spatial characteristics and the scale of aggregation. However, the standard linear model does not account for the different directions of scale-dependent proportion error since some classes become larger and others become smaller as the scene is aggregated. Addition of indicator variables representing class-type significantly improves the performance by allowing the model to respond differently to different classes. A regression tree model provides a much simpler fit to the complex scaling behavior through an interaction between patch size and aggregation scale. An understanding of the relationships between landscape pattern, scale, and proportion error may advance methods for correcting land-cover area estimates. Such methods could also facilitate high-resolution calibration and validation of coarse-scale remote-sensing-based land-cover mapping algorithms. Ongoing initiatives to produce global land-cover datasets from remote sensing, such as efforts within the IGBP and the EOS MODIS Land-Team, include significant emphasis on high level calibration and validation activities of this nature.

Twentieth-Century Droughts and Their Impacts on Terrestrial Carbon Cycling in China

Abstract Midlatitude regions experienced frequent droughts during the twentieth century, but their impacts on terrestrial carbon balance are unclear. This paper presents a century-scale study of drought effects on the carbon balance of terrestrial ecosystems in China. The authors first characterized the severe extended droughts over the period 1901–2002 using the Palmer drought severity index and then examined how these droughts affected the terrestrial carbon dynamics using tree-ring width chronologies and a process-based biogeochemistry model, the Terrestrial Ecosystem Model (TEM). It is found that China suffered from a series of severe extended droughts during the twentieth century. The major drought periods included 1920–30, 1939–47, 1956–58, 1960–63, 1965–68, 1978–80, and 1999–2002. Most droughts generally reduced net primary productivity (NPP) and net ecosystem productivity (NEP) in large parts of drought-affected areas. Moreover, some of the droughts substantially reduced the countrywide annual NPP...

Characteristics of composited AVHRR data and problems in their classification

Abstract Unsupervised classification procedures were applied to a temporal sequence of fifteen bi-weekly composited NDVI images produced from AVHRR LAC data. Individual examination of the input images appeared to show substantial contamination due to clouds which persist through the compositing period. These apparent cloud features dominated the results of the clustering procedures. The composites also include large numbers of far off-nadir pixels. This causes severe spatial smoothing and produces a blurred image appearance. Further combining the data to monthly composites largely eliminated the cloud cover problem, but did not necessarily reduce the frequency of large view zenith angles. Preprocessing of high temporal frequency; low spatial resolution data such as that provided by AVHRR and the planned EOS MODIS instrument must more effectively remove the effects of clouds, correct for anisotropic scattering from the atmosphere and bi-directional reflectance from the surface, and should be biased towards...

Geographical distribution of global greening trends and their climatic correlates: 1982-1998

We examined trends in vegetation activity at the global scale from 1982 to 1998 using a recently developed satellite‐based vegetation index in conjunction with a gridded global climate dataset. Vegetation greening trends were observed in the northern high latitudes, the northern middle latitudes, and parts of the tropics and subtropics. Temperature, and in particular spring warming, was the primary climatic factor associated with greening in the northern high latitudes and western Europe. Temperature trends also explained greening in the US Pacific Northwest, tropical and subtropical Africa, and eastern China. Precipitation was a strong correlate of greening in fragmented regions only. Decreases in greenness in southern South America, southern Africa, and central Australia were strongly correlated to both increases in temperature and decreases in precipitation. Over vast areas globally, strong positive trends in greenness exhibited no correlation with trends in either temperature or precipitation. These areas include the eastern United States, the African tropics and subtropics, most of the Indian subcontinent, and south‐east Asia. Thus, for large areas of land that are undergoing greening, there appears to be no climatic correlate. Globally, greening trends are a function of both climatic and non‐climatic factors, such as forest regrowth, CO2 enrichment, woody plant proliferation, and trends in agricultural practices.

Artificial neural network response to mixed pixels in coarse-resolution satellite data

Abstract A feedforward neural network model based on the multilayer perceptron structure and trained using the backpropagation algorithm responds to subpixel class composition in both simulated and real data. Maps of the network response surfaces for simulated data illustrate that the set of network outputs successfully characterizes the level of class dominance and the subpixel composition for controlled data that contain a range of class mixtures. For a Sierra Nevada test site, the correspondence between 250 m reference data and a network class map produced using 250 m degraded TM data depends on the degree of subpixed class mixing as determined from coregistered 30 m reference data. For most mislabeled pixels, classification error results from confusion between the first and second largest subpixel components, and the first and second largest network outputs. Overall map accuracy increases from 62% to 79% when mislabeled pixels are reclassified using the second largest network output. Accuracy increases to 84% if, for mislabeled pixels, the second largest subpixel class is used as a reference. Maps of the network response surfaces for a controlled subset of the Plumas data complement the findings of the simulated data and show that the network responds in a systematic way to changing proportions of subpixel components. Based on our results we suggest that interpretation of the complete set of network outputs can provide information on the relative proportions of subpixel classes. We outline a threshold-based heuristic that would allow the labeling of pure classes, mixed classes, and primary and secondary class types based on the relative magnitudes of the two largest network signals.

Trends in vegetation activity and their climatic correlates: China 1982 to 1998

We combined a satellite-derived Leaf Area Index (LAI) dataset and a gridded climate dataset to analyse trends in vegetation activity and their correlation with climate variability in China between 1982 and 1998. Vegetation activity over the growing season increased 11.03% in China during the 17-year period, which is broadly consistent with the greening trend in the northern high latitudes in Eurasia and North America shown in previous studies. Approximately 99×106 ha of croplands and 35×106 ha of forest exhibited significant upward trends in growing season LAI, and accounted for 53% and 19% of the total vegetated area with greening trends, respectively. Temperature was the leading climatic factor controlling greening patterns. However, trends in agricultural practices, such as increased use of high-yield crops and application of chemical fertilizers, along with land-use changes such as afforestation and reforestation probably made a greater contribution to the greening trend than temperature. Increased vegetation activity in forests suggests an increasing carbon stock in forest biomass in China, which supports previous studies based on satellite sensor data and forest inventory data.

A detail-preserving and flexible adaptive filter for speckle suppression in SAR imagery

Speckle filtering and detail preservation are two key issues in speckle suppression of Synthetic Aperture Radar (SAR) imagery. Different applications may require different balances between speckle reduction and detail retention. This paper presents a detail-preserving and flexible filter. Both quantitative and qualitative criteria, including speckle reduction, edge retention, texture preservation and visual assessment, were used to evaluate the performance of this adaptive filter. One JERS-1 SAR image and three SIR-C/X-SAR images (L-HH, L-HV and C-HV) were employed in the evaluation. The results show that the proposed filter is slightly better than, or comparable to commonly used filters based on the spatial domain, such as Lee, Frost, Lee-Sigma and Gamma-Map, in terms of detail preservation. Moreover, the proposed filter can achieve a wide range of balances between speckle reduction and detail preservation, and thus is applicable in different applications, including both broad-scale interpretation or mapping and applications in which fine details and high resolution are required. Furthermore, the proposed filter requires no knowledge of speckle standard deviation, which is required in most commonly used filters.

Automated mapping of conformity between topographic and geological surfaces

We present a technique to produce spatially distributed fields of geometric alignment between topography and the orientation of geologic bedding planes (topographic/bedding-plane intersection angle). Computation and digital mapping of the topographic/bedding-plane intersection angle (TOBIA) requires the derivation of four spatially distributed variables: topographic slope, slope aspect, bedding dip, and dip azimuth. Slope and slope aspect surfaces are derived from a high resolution (10 m) digital elevation model. Ordinary kriging is used to interpolate spatially continuous fields of dip azimuth and dip from point measurements of strike and dip. Using these variables, TOBIA can be mapped either categorically as slope types, or as a continuous index. Categorical mapping requires two steps. First, slopes are classified into three functional types based on the alignment between the dip azimuth and slope aspect. Slopes are then further partitioned based on the alignment between slope angle and dip angle. Continuous computations of TOBIA rely on a geometric equation using all four variables. The methods provide an eAcient means for estimating topographic/bedding plane intersection angles over large areas. Resulting surfaces are useful for a variety of landscape-scale modeling applications, such as the prediction of potential hillslope failure, hydrologic flow paths, and vegetation patterns. 7 2000 Elsevier Science Ltd. All rights reserved.

Multi-scale environmental heterogeneity as a predictor of plant species richness

Ecological theory predicts a positive influence of local-, landscape-, and regional-scale spatial environmental heterogeneity on local species richness. Therefore, knowing how heterogeneity measured at a variety of scales relates to local species richness has important implications for conservation of biological diversity. We took a statistical modeling approach to determine which metrics of heterogeneity measured at which scales were useful predictors of local species richness, and whether the heterogeneity-local richness relationship was always positive. Local plant species richness data came from 400-m2 vegetation plots in North and South Carolina, USA. At each of four scales from within plots to across regions, we used either GIS or field data to calculate measures of heterogeneity from abiotic environmental variables, vegetation productivity data, and land cover classifications. Among all predictors at all scales, we found that no measure of heterogeneity was a better predictor of local richness than mean pH within plots. However, at scales larger than within plots, measures of heterogeneity were correlated most strongly with local richness, and each of the three classes of variables we used had a distinct scale at which it performed better than the others. These results highlight the fact that ecological processes occurring across multiple scales influence local species richness differently. In addition, relationships between heterogeneity and richness were usually, though not always, positive, underscoring the importance of processes that occur at a variety of scales to local biodiversity conservation and management.