Year-end Sale % OFF 
Abstract
Remote-sensing tools and satellite data are often used to map and monitor changes in vegetation cover in forests and other perennial woody vegetation. Large-scale vegetation mapping from remote sensing is usually based on the classification of its spectral properties by means of spectral Vegetation Indices (VIs) and a set of rules that define the connection between them and vegetation cover. However, observations show that, across a gradient of precipitation, similar values of VI can be found for different levels of vegetation cover as a result of concurrent changes in the leaf density (Leaf Area Index—LAI) of plant canopies. Here we examine the three-way link between precipitation, vegetation cover, and LAI, with a focus on the dry range of precipitation in semi-arid to dry sub-humid zones, and propose a new and simple approach to delineate woody vegetation in these regions. By showing that the range of values of Normalized Difference Vegetation Index (NDVI) that represent woody vegetation changes along a gradient of precipitation, we propose a data-based dynamic lower threshold of NDVI that can be used to delineate woody vegetation from non-vegetated areas. This lower threshold changes with mean annual precipitation, ranging from less than 0.1 in semi-arid areas, to over 0.25 in mesic Mediterranean area. Validation results show that this precipitation-sensitive dynamic threshold provides a more accurate delineation of forests and other woody vegetation across the precipitation gradient, compared to the traditional constant threshold approach.
Highlights
Remote-sensing tools are widely used to map different attributes of vegetation at various spatial scales [1,2]
Wveegfuerttahteiron that we found in oshuorwanthaalty, assisaerxespullationf tthheedfiafflesree-nnceesgabetitvweeeanntdhefaculsreve-ps oofsitthievleowdeerliannedautipopnerotfhvreesghoeltdast,iothne in dry and range between the two thresholds of Normalized Difference Vegetation Index (NDVI) increases along the precipitation gradient
We used an empirical relation between NDVI and mean annual precipitation (MAP) and fitted a curve that can be used as a dynamic threshold for woody vegetation delineation
Summary
Remote-sensing tools are widely used to map different attributes of vegetation at various spatial scales [1,2]. Several methods are widely used for vegetation delineation: supervised classification (e.g., [5]), spectral methods (e.g., [6]), and machine-learning algorithms [7], or a simple constant lower threshold value (e.g., [8]) Vegetation mapping methods, such as classification, spectral unmixing and machine learning require large datasets for training, and require much computing power and time compared to the simple threshold approach and, are less efficient for large-scale mapping and long-term monitoring. We performed a qualitative analysis that illustrates the limitations of using VIs to monitor and interpret changes in vegetation cover and biomass To overcome these limitations, we developed a new simple method for woody vegetation delineation using a dynamic NDVI threshold as a function of MAP, based on a training set of NDVI and vegetation covers
Sign-in/Register to view highlights & summary Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a callDisclaimer: All third-party content on this website/platform is and will remain the property of their respective owners and is provided on "as is" basis without any warranties, express or implied. Use of third-party content does not indicate any affiliation, sponsorship with or endorsement by them. Any references to third-party content is to identify the corresponding services and shall be considered fair use under The CopyrightLaw.
