Seasonal Metrics Research Articles

Estimation of Mediterranean forest attributes by the application of k‐NN procedures to multitemporal Landsat ETM+ images

Routine applications of nonparametric estimation methods to satellite data for assisting the creation of forest inventories in Northern European countries are stimulating interest in the possible extension of these methods to more complex Mediterranean areas. This is the subject of the current work, which presents an experiment based on the integration of remotely sensed images and sample field measurements aimed at producing forest attribute maps in central Italy. Testing was carried out in an area where 370 geocoded field plots, sampled on a single‐stage cluster design, were collected to characterize wood and non‐wood forest attributes. These ground data served to apply various k‐Nearest Neighbour (k‐NN) estimation procedures to multitemporal Landsat 7 ETM+ images in order to map major forest attributes (basal area and simulated leaf area index, LAI). More specifically, the investigation focused on evaluating the effects of using satellite images from different periods of the growing season and spectral metrics of increasing complexity. The results achieved by the examined methods are finally discussed in order to provide guidelines for possible operational utilization.

An analysis of relationships among plant community phenology and seasonal metrics of Normalized Difference Vegetation Index in the northern part of the monsoon region of China.

This study focuses on relationships between the phenological growing season of plant communities and the seasonal metrics of Normalized Difference Vegetation Index (NDVI) at sample stations and pixels overlying them, and explores the procedure for determining the growing season of terrestrial vegetation at the regional scale, using threshold NDVI values obtained by surface-satellite analysis at individual stations/pixels. The cumulative frequency of phenophases has been calculated for each plant community and each year in order to determine the growing season at the three sample stations from 1982 to 1993. The precise thresholds were arbitrarily set as the dates on which the phenological cumulative frequency reached 5% and 10% (for the beginning) and 90% and 95% (for the end). The beginning and end dates of the growing season were then applied each year as time thresholds, to determine the corresponding 10-day peak greenness values from NDVI curves for 8-km2 pixels overlying the phenological stations. According to a trend analysis, a lengthening of the growing seasons and an increase of the integrated growing season NDVI have been detected in the central part of the research region. The correlation between the beginning dates of the growing season and the corresponding threshold NDVI values is very low, which indicates that the satellite-sensor-derived greenness is independent of the beginning time of the growing season of local plant communities. Other than in spring, the correlation between the end dates of the growing season and the corresponding threshold NDVI values is highly significant. The negative correlation shows that the earlier the growing season terminates, the larger the corresponding threshold NDVI value, and vice versa. In order to estimate the beginning and end dates of the growing season using the threshold NDVI values at sites without phenological data from 1982 to 1993, we calculated the spatial correlation coefficients between NDVI time-series at each sample station and other contiguous sites year by year. The results provide the spatial extrapolation area of the growing season for each sample station. Thus, we can use the threshold NDVI value obtained at one sample station/pixel for a year to determine the growing season at the extrapolation sites with a similar vegetation type for the same year.

Global land cover classifications at 8 km spatial resolution: The use of training data derived from Landsat imagery in decision tree classifiers

This paper reports a study which aims to (i) develop methodologies for global land cover classifications that are objective, reproducible and feasible to implement as new satellite data become available in the future and (ii) provide a global land cover classification product based on the National Aeronautics and Space Administration/National Oceanic and Atmospheric Administration Pathfinder Land (PAL) data that can be used in global change research. The spatial resolution for the land cover classification is 8 km, intermediate between our previously published coarse one degree by one degree spatial resolution and the 1 km global land cover product being developed under the auspices of the International Geosphere Biosphere Program. We first derive a global network of training sites from Landsat imagery, using 156 Landsat scenes mostly from the Multispectral Scanner System, to identify over 9000 pixels in the PAL data where we have high confidence that the labelled cover type occurs. We then use the training data to test a number of metrics that describe the temporal dynamics of vegetation over an annual cycle for potential use as input variables to a global land cover classification. The tested metrics are based on: (i) the ratio between surface temperature and Normalized Difference Vegetation Index (NDVI); (ii) seasonal metrics derived from the NDVI temporal profile, such as length of growing season; (iii) a rule-based approach that determines cover type through a series of hierarchical trees based on surface temperature and NDVI values; and (iv) annual mean, maximum, minimum and amplitude values for all optical and thermal channels in the Advanced Very High Resolution Radiometer (AVHRR) (PAL) data. Highest mean class accuracies from a decision tree classifier were obtained using the annual mean, maximum, minimum, and amplitude values for all AVHRR bands. Finally, we apply these metrics to 1984 PAL data at 8 km resolution to derive a global land cover classification product using a decision tree classifier. The classification has an overall accuracy between 81.4 and 90.3%. The Landsat images used for deriving the training data and the methodology for classification of AVHRR data at 8km resolution can also be applied to 1 km AVHRR data and, in the future, Moderate Resolution Imaging Spectroradiometer (MODIS) data at 250 and 500m resolution. Digital versions of the land cover dataset and detailed documentation can be found on the World Wide Web at http://www.geog.umd.edu/landcover/8km-map.html.