Plant phenology is the study of the timing of important life cycle events, and how they relate to variations in climatic conditions. The study of plant phenology has gained significant interest recently with the pressing threat of climate change. Changes in phenology have long been regarded as an important and sensitive indicator of changes in climate. Because spring leafout and blooming times are very closely linked to preceding temperatures, changes in phenology are one of the easiest to document and most immediate impacts of climate change on ecosystems (Schwartz et al. 2006). In temperate zones across the planet, the timing of phenological events has been shifting, and this has been directly linked to recent warming (e.g. Menzel et al. 2006; Beaubien and Hamann 2011a). Phenology determines the length of the growing season, which is the time between spring leafout and fall senescence. This is of vital interest to climate science because growing season length is a major determinant of carbon uptake in an ecosystem, and thus its ability to mitigate against climate change through carbon sequestration (Cleland et al. 2007).
Citizen science is a useful way to collect phenological data, particularly in areas where there are many motivated and knowledgeable volunteers who can observe and collect this information. In 1987, Elisabeth Beaubien started the Plant Watch citizen science network to collect phenology data in Alberta, Canada. The study region is more northerly and cooler than any of the other regions in North American for which there are extensive phenology records (Bertin 2008). As of 2016, the PlantWatch database included phenology observations for 30 species, with more than 57,000 individual observations from roughly 700 observers. These data were combined with two other sets of phenological records to show that from 1936 to 2006 bloom dates for the two earliest blooming species had become significantly earlier (Anemone patens: -2.1 days/decade; Populus tremuloides: -2.0 days/decade); this corresponds to a shift of two weeks over the study period (Beaubien and Hamann 2011a).
Remote sensing has recently become a valuable tool to complement ground collected phenology data. Remote sensing strategies are generally used to measure green-up, also termed as "land surface phenology", which is when the satellite pixel begins to become green or reach a certain percentage of full summer greenness (Delbart et al. 2015). Good satellite records exist since the early 1980s, and can provide data on the earth's surface frequently enough to document the gradual changes in surface reflectance that correspond to plants leafing out and flowering in spring. Remotely sensed phenology has the advantage of being able to document phenology over a greater spatial extent, and provide records for areas where there are few or no potential observers. However, remote sensing measurements of phenology can be biased or erroneous due to particular land cover types, such as cropland, waterbodies, or urban areas (Delbart et al. 2015). For this reason, it is useful to compare and calibrate land surface phenology records with leafing and flowering data collected on the ground.
The objective of this study was to evaluate how land surface phenology greenup dates relate to flower bloom times and leaf-out dates of Alberta plants between 1987-2016. We test how these data correspond to one another based on landcover types for the best represented ecoregions within the ground phenological observations. We expected that the PlantWatch data and land surface phenology data would have high correlations for natural landscapes (i.e. grasslands, forests), and low to no correlation for non-natural or heavily modified landscapes (i.e. cropland).
Citizen science is a useful way to collect phenological data, particularly in areas where there are many motivated and knowledgeable volunteers who can observe and collect this information. In 1987, Elisabeth Beaubien started the Plant Watch citizen science network to collect phenology data in Alberta, Canada. The study region is more northerly and cooler than any of the other regions in North American for which there are extensive phenology records (Bertin 2008). As of 2016, the PlantWatch database included phenology observations for 30 species, with more than 57,000 individual observations from roughly 700 observers. These data were combined with two other sets of phenological records to show that from 1936 to 2006 bloom dates for the two earliest blooming species had become significantly earlier (Anemone patens: -2.1 days/decade; Populus tremuloides: -2.0 days/decade); this corresponds to a shift of two weeks over the study period (Beaubien and Hamann 2011a).
Remote sensing has recently become a valuable tool to complement ground collected phenology data. Remote sensing strategies are generally used to measure green-up, also termed as "land surface phenology", which is when the satellite pixel begins to become green or reach a certain percentage of full summer greenness (Delbart et al. 2015). Good satellite records exist since the early 1980s, and can provide data on the earth's surface frequently enough to document the gradual changes in surface reflectance that correspond to plants leafing out and flowering in spring. Remotely sensed phenology has the advantage of being able to document phenology over a greater spatial extent, and provide records for areas where there are few or no potential observers. However, remote sensing measurements of phenology can be biased or erroneous due to particular land cover types, such as cropland, waterbodies, or urban areas (Delbart et al. 2015). For this reason, it is useful to compare and calibrate land surface phenology records with leafing and flowering data collected on the ground.
The objective of this study was to evaluate how land surface phenology greenup dates relate to flower bloom times and leaf-out dates of Alberta plants between 1987-2016. We test how these data correspond to one another based on landcover types for the best represented ecoregions within the ground phenological observations. We expected that the PlantWatch data and land surface phenology data would have high correlations for natural landscapes (i.e. grasslands, forests), and low to no correlation for non-natural or heavily modified landscapes (i.e. cropland).