Differentiating plant species within and across diverse ecosystems with imaging spectroscopy

Roth, Keely L.; Roberts, Dar A.; Dennison, Philip E.; Alonzo, Michael; Peterson, S.; Beland, Michael

doi:10.1016/j.rse.2015.05.007

Cited by 82 publications

(70 citation statements)

References 76 publications

Supporting

Mentioning

Contrasting

Unclassified

Order By: Relevance

“…Mortality should result in more persistent reductions in RGVF, however, recruitment of herbaceous vegetation following mortality could introduce more seasonal variability. Hyperspectral VSWIR data have demonstrated the ability to map dominant plant species [19][20][21][22][23][24]30], which could be used to separate senescence and dieback (persistence of the same species) from mortality and replacement of one dominant species with another. Dominant species cover mapped using hyperspectral data could be used to extend these results beyond a limited number of field-assessed reference polygons.…”

Section: Resultsmentioning

confidence: 99%

“…Reference polygons were required to be at least 75% dominated by a single species (or in one case by intermixed Artemisia californica and Salvia leucophylla) and to have a minimum approximate size of 40 m by 40 m. Reference polygons were mapped within the study area in 2003, 2009, and 2012 ( Figure 1), and were collected from outside of areas burned by recent fires in the Santa Ynez Mountains. Reference data collection is described in further detail in [21] and [41].…”

Section: Ground Reference Datamentioning

confidence: 99%

“…Hyperspectral data have capabilities for estimating vegetation biophysical and biochemical parameters [15][16][17][18] and mapping vegetation species [19][20][21][22][23][24]. These complementary capabilities give hyperspectral data strong potential for monitoring drought impacts on vegetation [25,26], but a paucity of hyperspectral time series datasets has prevented the development of drought monitoring applications.…”

Section: Introductionmentioning

confidence: 99%

See 2 more Smart Citations

Monitoring the Impacts of Severe Drought on Southern California Chaparral Species using Hyperspectral and Thermal Infrared Imagery

et al. 2015

View full text Add to dashboard Cite

Airborne hyperspectral and thermal infrared imagery acquired in 2013 and 2014, the second and third years of a severe drought in California, were used to assess drought impacts on dominant plant species. A relative green vegetation fraction (RGVF) calculated from 2013-2014 Airborne Visible Infrared Imaging Spectrometer (AVIRIS) data using linear spectral unmixing revealed seasonal and multi-year changes relative to a pre-drought 2011 reference AVIRIS image. Deeply rooted tree species and tree species found in mesic areas showed the least change in RGVF. Coastal sage scrub species demonstrated the highest seasonal variability, as well as a longer-term decline in RGVF. Ceanothus species were apparently least well-adapted to long-term drought among chaparral species, showing persistent declines in RGVF over 2013 and 2014. Declining RGVF was associated with higher land surface temperature retrieved from MODIS-ASTER Airborne Simulator (MASTER) data. Combined collection of hyperspectral and thermal infrared imagery may offer new opportunities for mapping and monitoring drought impacts on ecosystems.

show abstract

Section: Resultsmentioning

confidence: 99%

Section: Ground Reference Datamentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

Monitoring the Impacts of Severe Drought on Southern California Chaparral Species using Hyperspectral and Thermal Infrared Imagery

et al. 2015

View full text Add to dashboard Cite

show abstract

“…One potential way to avoid the reduction in overall accuracy, is to adapt available methods to reduce the size of large species classes while maintaining the full range of spectral variaiblity in that class. This has been done for species classification using spectral mixture analysis and discriminant analysis [26,53]. For SVM algorithms, spectral variability can be maintined while reducing large data classes with numerous methods to balance the data [49], by choosing only data points on the border between classes to define the separation between classes [22], or iteratively pruning the support vectors to achieve the best separation between classes [54].…”

Section: Selection Of Training Data For Optimal Species-level Accuracymentioning

confidence: 99%

Tree Species Abundance Predictions in a Tropical Agricultural Landscape with a Supervised Classification Model and Imbalanced Data

et al. 2016

View full text Add to dashboard Cite

Abstract:Mapping species through classification of imaging spectroscopy data is facilitating research to understand tree species distributions at increasingly greater spatial scales. Classification requires a dataset of field observations matched to the image, which will often reflect natural species distributions, resulting in an imbalanced dataset with many samples for common species and few samples for less common species. Despite the high prevalence of imbalanced datasets in multiclass species predictions, the effect on species prediction accuracy and landscape species abundance has not yet been quantified. First, we trained and assessed the accuracy of a support vector machine (SVM) model with a highly imbalanced dataset of 20 tropical species and one mixed-species class of 24 species identified in a hyperspectral image mosaic (350-2500 nm) of Panamanian farmland and secondary forest fragments. The model, with an overall accuracy of 62%˘2.3% and F-score of 59%˘2.7%, was applied to the full image mosaic (23,000 ha at a 2-m resolution) to produce a species prediction map, which suggested that this tropical agricultural landscape is more diverse than what has been presented in field-based studies. Second, we quantified the effect of class imbalance on model accuracy. Model assessment showed a trend where species with more samples were consistently over predicted while species with fewer samples were under predicted. Standardizing sample size reduced model accuracy, but also reduced the level of species over-and under-prediction. This study advances operational species mapping of diverse tropical landscapes by detailing the effect of imbalanced data on classification accuracy and providing estimates of tree species abundance in an agricultural landscape. Species maps using data and methods presented here can be used in landscape analyses of species distributions to understand human or environmental effects, in addition to focusing conservation efforts in areas with high tree cover and diversity.

show abstract

“…Spectra are thus an aggregate signal of the chemical and structural composition of vegetation, and can be directly related to a number of leaf biochemical and morphological functional traits (Table 1; [30][31][32]. Air-or satellite-borne spectrometers are able to measure the aggregate functional traits of plant communities represented in the top layers of vegetation, and even the attributes of single species directly, depending on community spatial and spectral characteristics 33 . This capability has been successfully demonstrated using airborne spectrometers for many traits at regional scales across multiple biomes 34,35 .…”

mentioning

confidence: 99%

Erratum: Monitoring plant functional diversity from space

Jetz¹,

Cavender‐Bares²,

Pavlick³

et al. 2016

Nature Plants

View full text Add to dashboard Cite

The world's ecosystems are losing biodiversity fast. A satellite mission designed to track changes in plant functional diversity around the globe could deepen our understanding of the pace and consequences of this change and how to manage it.The ability to view Earths' vegetation from space is a hallmark of the space age. Yet decades of satellite measurements have provided relatively little insight into the immense diversity of form and function in the plant kingdom in space and time. Humans are rapidly impacting biodiversity around the globe 1,2 , leading to the loss of ecosystem function 3 , and the goods and services they provide 4,5 . Recognizing the gravity of this threat, the international community has committed to urgent action to halt biodiversity loss [6][7][8][9] .Ecosystem processes [10][11][12] are often directly linked to the functional biodiversity of plants, that is, to a wide range of plant chemical, physiological and structural properties, connected to the uptake, use and allocation of resources. The functional biodiversity of plants varies in space and time and across scales of biological organization. Capturing and understanding this variation is vitally important for tracking the status and resilience of Earth's ecosystems, and for predicting how our ecological life support systems will function in the future.We currently lack consistent, repeated, high-resolution global-scale data on the functional biodiversity of the Earth's vegetation 2,10-12 . However, the technological tools, informatics infrastructure, theoretical basis, and analytical capability now exist to produce this essential data. Here we suggest that this capability is utilized in a satellite mission supporting a Global Biodiversity Observatory, that tracks temporal changes in plant functional traits across the globe to fill critical knowledge gaps, aid in the assessment of global environmental change, and improve predictions of future change. The continuous, global coverage in space and time such a mission would provide has the potential to transform basic and applied science on diversity and function, and to pave the way to a more mechanistically detailed representation of the terrestrial biosphere in Earth system models.The data and knowledge gap Plant functional biodiversity encompasses the wide-ranging variation in the chemical physiological and morphological properties of plants, such as the concentration of metabolites and nonstructural carbohydrates in leaves and the ratio of leaf mass to leaf area. These attributes are related functionally to the uptake, allocation and use of resources, such as carbon and nutrients, within the plant, and to defense against pests and environmental stresses.Functional properties vary within and among individuals (for instance as determined by the position of a leaf on a plant, or a tree in a forest), populations, species and communities, and may be measured at any of these levels of biological organization. With increasing spatial scale (and thus decreasing spatial resolution of measurem...

show abstract

Differentiating plant species within and across diverse ecosystems with imaging spectroscopy

Cited by 82 publications

References 76 publications

Monitoring the Impacts of Severe Drought on Southern California Chaparral Species using Hyperspectral and Thermal Infrared Imagery

Monitoring the Impacts of Severe Drought on Southern California Chaparral Species using Hyperspectral and Thermal Infrared Imagery

Tree Species Abundance Predictions in a Tropical Agricultural Landscape with a Supervised Classification Model and Imbalanced Data

Erratum: Monitoring plant functional diversity from space

Contact Info

Product

Resources

About