Extensive Green Roof Species Evaluations Using Digital Image Analysis

Abstract

<jats:p>This research examined plant area covered (plant cover) for six plant species on an existing modular extensive green roof in a semiarid climate. Species evaluated were <jats:italic>Antennaria parvifolia</jats:italic> Nutt., <jats:italic>Bouteloua gracilis</jats:italic> (Kunth) Lag., <jats:italic>Delosperma cooperi</jats:italic> (Hook. f.) L. Bol., <jats:italic>Eriogonum umbellatum</jats:italic> Torr. <jats:italic>aureum</jats:italic> ‘Psdowns’, <jats:italic>Opuntia fragilis</jats:italic> Nutt., and <jats:italic>Sedum lanceolatum</jats:italic> Torr. Many methods for measuring plant cover are subjective and based on semiquantitative measurements. This study compared digital image analysis data (DIA) with manually collected converted two-dimensional data (C2D) for plants grown on an extensive green roof. For each plant in the study, digital images and manual two-dimensional measurements were taken on four dates (at 6-week intervals) in 2008 and on four dates (at 6-week intervals) in 2009. Additionally, comparisons between DIA data and final biomass, and C2D and final biomass, were performed. Plant cover increased for all six species during the 2008 growing season. However, <jats:italic>E. umbellatum aureum</jats:italic> ‘Psdowns’ had a low overwintering rate (12.5%) and was removed from analysis in 2009. In the spring of 2009, four of the five remaining species exhibited decreased plant cover as a result of winter dieback; the one exception was <jats:italic>O. fragilis</jats:italic>. In terms of plant cover, both quantification methods (C2D and DIA) revealed that <jats:italic>B. gracilis</jats:italic> and <jats:italic>D. cooperi</jats:italic> had more plant cover than <jats:italic>A. parvifolia</jats:italic>, <jats:italic>O. fragilis</jats:italic>, and <jats:italic>S. lanceolatum</jats:italic> by the end of the study. Thus, five of the six species evaluated in this study are appropriate for use in extensive green roof applications in semiarid regions. High levels of correlation were found between the DIA and C2D data sets (<jats:italic>r</jats:italic> = 0.77) averaged over the five species on all eight data collection dates. The groundcover species (<jats:italic>A. parvifolia</jats:italic>, <jats:italic>D. cooperi</jats:italic>, and <jats:italic>S. lanceolatum</jats:italic>) had a higher correlation on average (<jats:italic>r</jats:italic> = 0.83) than the upright (<jats:italic>B. gracilis</jats:italic>, <jats:italic>r</jats:italic> = 0.70) and decumbent (<jats:italic>O. fragilis</jats:italic>, <jats:italic>r</jats:italic> = 0.65) species. Additionally, DIA and final biomass correlations showed parallel trends with groundcovers averaging <jats:italic>r</jats:italic> = 0.83, upright <jats:italic>r</jats:italic> = 0.64, and decumbent <jats:italic>r</jats:italic> = 0.41. Therefore, using DIA to evaluate plant cover and biomass accumulation is especially appropriate for groundcover species.</jats:p>

Solar Green Roofs: A Unified Outlook 20 Years On

“…The green roof areas beneath and close to the PV array remained fully vegetated, whereas coverage in the exposed areas was patchy which is, the scholars noted, a typical condition in moisture‐deprived plants exposed to high levels of direct solar irradiance (bottom of Figure ) …”

“…Testing a properly installed set of PV modules over a green roof comprising eight plant species including Opuntia fragilis (Figure ), a team of scholars in Colorado found that the shade cast by the PV module string during the morning hours establishes gradients in substrate moisture, solar irradiance, and temperatures, with the substrate surface acting as a thermal sink relative to exposed roof surfaces . The temperature variations below the PV modules, for example, were markedly lower amounting to 3.92 and 6.36 °C, for the substrate surface beneath the modules in sheltered areas and in exposed areas, respectively.…”

“…Effects of protection from the PV array on the green roof (top) where vegetation appears to have greater coverage (bottom left) in areas shaded from direct sunlight for a portion of each day (bottom right).Reproduced with permission . Copyright 2017, Green Infrastructure Foundation. …”

Influence of vegetation damage on urban cooling effects

“…This provides a measure of the interface between canopy and atmosphere but can be difficult to quantify due to spatial (horizontal and vertical) and temporal variability. On a green roof, vegetation cover is a useful indicator of green roof health and can be estimated via digital analysis of photographic images (Bousselot et al, 2010). Damage can be assessed by comparing measured LAIs with those obtained from the same species grown in an optimum or stress-free environment.…”

Performance of Zoysia matrella ‘Zeon’ in Shallow Green Roof Substrates under Moisture Deficit Conditions

“…Each image was cropped using Adobe Photo-Shop CS3 software (Adobe Systems Inc., CA) to remove any unwanted image portion. Green Turf Cover percentage was determined using digital image analysis SigmaScan Pro Version 5.0 software (SigmaScan Pro; Systat Software Inc., Chicago, IL) (Bousselot et al, 2010;Karcher and Richardson, 2005;Olmstead et al, 2004;Richardson et al, 2001). Green leaves were selectively identified in the images by setting the hue range from 47 to 107 and saturation from 0 to 100.…”