you can find this page online at www.usanpn.org/
skip to content

USA National Phenology Network

Phenology: the Pulse of Our Planet.

Remote Sensing

Contributions of Remote Sensing to USA-NPN

The remote sensing component of the National Phenology Network is critical for providing a means of scaling from on-the-ground observations of phenology to the ecosystem level and, ultimately, to wall-to-wall estimates of phenology covering the entire country.  If this scaling is successful, it also provides a means to go back in time, using satellite observations dating back to 1989 (at 1km ground resolution), thus providing a longer time-series of phenology measures from which to establish short- and medium-term trends.

Remote Sensing Data for Phenological Studies

Applications of Landsat data paved the way for remote sensing-based phenology and these data are still used for some applications.  However, current research using satellite sensors with a more frequent repeat cycle currently dominate the field.  The longest-running series of high repeat-frequency sensors is the National Oceanic and Atmospheric Administration’s (NOAA) advanced very high resolution radiometer (AVHRR).  The AVHRR series has a near daily repeat cycle of the Earth and a 1km spatial resolution.  Both the temporal density of the data and the moderate spatial resolution make this sensor well suited for studying large area phenology.  AVHRR vegetation index data are available in a consistently processed database from 1982 at an 8km resampling grid covering the globe (http://glcf.umiacs.umd.edu/data/gimms/) and at 1km resolution from 1989 covering the conterminous United States (http://edcsns17.cr.usgs.gov/EarthExplorer/). 

More recent and much better calibrated moderate resolution satellite sensors that have the proper instrumentation for studying vegetation greenness include SPOT Vegetation (1-km data, launched in 1998) and Envisat MERIS (300-m data, launched in 2002).  The moderate resolution imaging spectroradiometer (MODIS) launched in December, 1999, and MODIS is frequently used for phenology studies.  Improved geometry, radiometry, and overall data quality of MODIS, combined with its free-of-charge data policy, provide readily available high quality data for time-series analysis.  One of the standard MODIS land products is the MOD12Q2 product (Land Cover Dynamics), distributed from the Land Processes Distributed Active Archive Center, which includes phenological data.

The key phenological variables that are often estimated from satellite remote sensing are the start, end, and length of the growing season.  These metrics are derived using algorithms that are based on a number of strategies, including threshold values, inflection points in time-series greenness curves, or rates of change in vegetation index values.
The seasonal pattern of variation in vegetated land surfaces that can be observed from remote sensing can be described by the term land surface phenology.  While the observed patterns are certainly related to biological phenomena, land surface phenology is distinct from traditional definitions of vegetation phenology, which refer to specific life cycle events such as budbreak, flowering, or leaf senescence using in situ observations of individual plants or species.  Although the meaning of land surface phenology in many ecosystems is clear, there are many environments in which the precise meaning is less clear such as mixed forests, evergreen forests and drylands.

USA-NPN: Linking Land Surface and Remote Sensing Observations

Land surface phenology, coupled with USA-NPN ground-based observations, provide a potential to progress from plant-specific estimates to characterizing land surface phenology on a wall-to-wall national scale.  But scaling from point observations to the landscape scale is a challenging research area.  The required set of in situ measurements is presently not well-defined and conventional field measurements consist of species-specific observations of canopy properties, which are not sufficient for characterizing land surface phenology.  Coupling USA-NPN observations with cooperating ecological networks (e.g., LTER, Ameriflux, National Park Service) provides a valuable opportunity for connecting ground observations to the landscape scale, provided specialized observation strategies are followed.

USA-NPN Remote Sensing Strategy and Activities

With these issues in mind, the remote sensing group of NPN constructed short- and medium-term strategies for implementing the remote sensing component of NPN.  The short-term strategy includes plans for both in situ validation/characterization pilot studies of LSP at intensive observation sites (in cooperation with several LTERs) and an observation protocol for cooperating extensive observation sites that will make broader-scale observations.  The pilot experiment will involve collecting weekly measurements (including digital photographs) of community greenness at fixed transect sites that can be compared to Landsat and MODIS observed fractional greenness.  The extensive observations will provide information on the spatial patterns of stages of vegetation community phenology that can be compared to both remote sensing and species level observations to forge a stronger link between the two approaches.  The mid-term goals include development of a web site for the group, initiation of a field validation study of the intensive observation sites (developed from lessons learned during the pilot study), and development of proposals involving validation techniques, algorithm intercomparison studies, and forecast model development. 

A wealth of remote-sensing research fields will be opened with the availability of the NPN data sets, including: 1) comparative studies of alternative phenology characterization techniques; 2) techniques to validate remote sensing phenology parameters; 3) characterization of relationships of key phenology parameters to climatic variables; 4) characterization of relationships of remote sensing-based phenology parameters to indicator species and natural vegetation; and 5) scaling studies.  Phenological data collected by the various tiers of the NPN will be a critical source for validating land surface phenology estimates from satellite sensors and, conversely, estimates from remote sensing can fill gaps between ground observations to produce a continuous surface of phenology estimates at the continental scale.

MODIS NDVI CAYO Reed 2005MODIS image of phenology in the vicinity of Canyonlands National Park, with
on-ground measurement sites indicated by crosses (image courtesy of Dr.
Bradley Reed, USGS).

 

back to top