The effect of seasonal spectral variation on species classification in the Panamanian tropical forest

Plant species identification and mapping based on remotely-sensed spectral signatures is a challenging task with the potential to contribute enormously to ecological studies. Success in this task rests upon the appropriate collection and use of costly field-based training data, and researchers are in need of ways to improve collection efficiency based on quantitative evidence. Using imaging spectrometer data collected by the Carnegie Airborne Observatory for hundreds of field-identified tree crowns in Kruger National Park, South Africa, we developed woody plant species classification models and evaluated how classification accuracy increases with increasing numbers of training crowns. First, we show that classification accuracy must be estimated while respecting the crown as the basic unit of data; otherwise, accuracy will be overestimated and the amount of training data needed to perform successful classification will be underestimated. We found that classification accuracy and the number of training crowns needed to perform successful classification varied depending on the number and spectral separability of species in the model. We also used a modified Michaelis-Menten function to describe the empirical relationship between training crowns and model accuracy, and show how this function may be useful for predicting accuracy. This framework can assist researchers in designing field campaigns to maximize the efficiency of field data collection, and thus the amount of biodiversity information gained from remote species identification models.

Section: Discussionmentioning

confidence: 83%

Improving Remote Species Identification through Efficient Training Data Collection

Baldeck

Asner

2014

“…Although clear spectral differences among the major groups of species were demonstrated, previous findings [2] were based on a single January 2005 image acquisition. Seasonal and inter-seasonal variation in canopy spectral properties were ignored, yet phenological differences throughout and between growing seasons can affect the spectral separability of species [6,7]. A temporal analysis of the spectral separability is thus important not only to broaden our understanding of how to translate the biochemical and structural properties of the vegetation to species composition, but also to place previous results in the proper temporal context.…”

Section: Introductionmentioning

confidence: 99%

Hyperspectral Time Series Analysis of Native and Invasive Species in Hawaiian Rainforests

Somers

Asner

2012

Abstract:The unique ecosystems of the Hawaiian Islands are progressively being threatened following the introduction of exotic species. Operational implementation of remote sensing for the detection, mapping and monitoring of these biological invasions is currently hampered by a lack of knowledge on the spectral separability between native and invasive species. We used spaceborne imaging spectroscopy to analyze the seasonal dynamics of the canopy hyperspectral reflectance properties of four tree species: (i) Metrosideros polymorpha, a keystone native Hawaiian species; (ii) Acacia koa, a native Hawaiian nitrogen fixer; (iii) the highly invasive Psidium cattleianum; and (iv) Morella faya, a highly invasive nitrogen fixer. The species specific separability of the reflectance and derivative-reflectance signatures extracted from an Earth Observing-1 Hyperion time series, composed of 22 cloud-free images spanning a period of four years and was quantitatively evaluated using the Separability Index (SI). The analysis revealed that the Hawaiian native trees were universally unique from the invasive trees in their near-infrared-1 (700-1,250 nm) reflectance (0.4 > SI > 1.4). Due to its higher leaf area index, invasive trees generally had a higher near-infrared reflectance. To a lesser extent, it could also be demonstrated that nitrogen-fixing trees were spectrally unique from non-fixing trees. The higher leaf nitrogen content of nitrogen-fixing trees was expressed through slightly increased separabilities in visible and shortwave-infrared reflectance wavebands (SI = 0.4). We also found phenology to be key to spectral separability analysis. As such, it was shown that the spectral separability in the near-infrared-1 reflectance OPEN ACCESSRemote Sens. 2012, 4 2511 between the native and invasive species groups was more expressed in summer (SI > 0.7) than in winter (SI < 0.7). The lowest separability was observed for March-July (SI < 0.3). This could be explained by the invasives taking advantage of the warmer summer period to expand their canopy. There was, however, no specific time window or a single spectral region that always defined the separability of all species groups, and thus intensive monitoring of plant phenology as well as the use of the full-range (400-2,500 nm) spectrum was highly advantageous in differentiating each species. These results set a basis for an operational invasive species monitoring program in Hawai'i using spaceborne imaging spectroscopy.

“…Depending on specific tree species, the phenology information is very useful and often used in tree species classification [35]. For example, rubber is a deciduous tree species, and integration of different seasons of optical sensor data has been proven valuable in improving rubber extraction accuracy [13,16,19].…”

Section: Introductionmentioning

confidence: 99%

Mapping Torreya grandis Spatial Distribution Using High Spatial Resolution Satellite Imagery with the Expert Rules-Based Approach

Wang

2017

Rapid expansion of Torreya forests in the mountainous region in Zhejiang Province in the past three decades has produced many environmental problems such as soil erosion and poor water quality, requiring an update of its spatial distribution in a timely way. However, to date there are no suitable approaches available for mapping Torreya forest distribution, especially the new Torreya plantations, due to the complex landscapes. This research used high spatial resolution Chinese Gaofen (GF-1) and Ziyuan (ZY-3) satellite images and digital elevation model (DEM) data to extract old Torreya forests and new Torreya plantations using a newly proposed expert rules-based approach. Different variables such as spectral bands, vegetation indices, textural images, and DEM-derived variables were examined, and separability analyses of different land covers were explored. An expert rules-based approach was developed for the extraction of old Torreya forests and new Torreya plantations. The accuracy assessment using field survey data and Google Earth images indicates that this newly-proposed approach can effectively distinguish both old Torreya forests and new Torreya plantations from other land covers with producer's accuracies of 84% and 92%, and user's accuracies of 77% and 85%, respectively, much better classification accuracies than the maximum likelihood classifier. This new approach may be used for other study area for extracting Torreya forest distribution. This research provides valuable data sources for better managing existing Torreya forests and planning potential Torreya expansions in this region in the near future.