• Home   /  
  • Archive by category "1"

Lhermitte Evaluation Essay

1. Introduction

Fire regimes are characterized by their spatial pattern, frequency, intensity, seasonality, size distribution and severity [1]. In recent years, measurements of severity have gained importance. Severity is a valuable substitute for fire intensity when data on fire intensity are unavailable. Fire intensity describes the physical combustion process in terms of energy release from organic matter [2]. As a result, fire intensity is generally expressed in energy fluxes. Severity, in contrast, is more general in gauging the fire impact. This impact can be described as: (i) the amount of damage [3,4,5]; (ii) the physical, chemical and biological changes [6,7,8,9,10]; or (iii) the degree of alteration [11,12] that fire causes to an ecosystem. In this context, the terms fire severity and burn severity are often used interchangeably [2], however, Lentile et al. [13] and Veraverbeke et al. [14], suggest a clear distinction between both terms by considering the fire disturbance continuum [15], which addresses three different temporal fire effects phases: before, during and after the fire. In this framework fire severity quantifies the short-term fire effects in the immediate post-fire environment while burn severity quantifies both the short- and long-term impact as it includes response processes. A precise assessment of fire/burn severity is essential to: (i) obtain more reliable estimates of burning efficiency, which is a crucial parameter for evaluating greenhouse gas emissions [16]; and to (ii) contribute to a better understanding of fire regimes, regenerative strategies of species, successional pathways and hydro-geomorphological effects. These important objectives strengthen the need for a better understanding of fire/burn severity as an integral component in ecosystem functioning.

From a mono- and bi-temporal point of view conceptually simple band ratioing as well as more sophisticated approaches, such as spectral unmixing and radiative transfer models (RTMs), have been used to estimate wildfire severity, traditionally with moderate resolution Landsat imagery. RTMs consider the whole spectral profile and are physically based [17,18,19,20]. The main advantage of these simulation models is that their performance is site-independent which greatly enhances their applicability and inter-comparability over a wide range of ecosystems [18,20]. Spectral mixture analysis (SMA) applied to post-fire images have resulted in fractional ground cover measures closely related to burning efficiency, usually implementing at least the green vegetation and charred soil endmembers [21,22,23]. SMA proved to be efficient in detecting the charcoal signal even in lightly burned areas that kept a strong vegetation signal. Unmixing is therefore advantageous because of its ability to distinguish between burns and other sparsely vegetated areas [24]. The most popular approach, however, can be found in ratioing band reflectance data. In this context, the Normalized Burn Ratio (NBR) has become accepted as the standard spectral index to assess the severity of fire [25,26,27,28,29]. The NBR relates to vegetation moisture content by combining near infrared (NIR) with short wave infrared (SWIR) reflectance. Pre-and post-fire ratio images are often bi-temporally differenced, resulting in the differenced layers, which permit a clear distinction between burned and unburned region [27]. The bi-temporal index approach, however, can be constrained by limitations in image availability [28], which reduces the chances for acquiring the ideal anniversary date image couple [29,30]. For this reason, extracting critical post-fire information from a single scene can be of particular interest.

Previous research has demonstrated that the performance of the NBR approach varies among ecoregions and vegetation types [26,27,31,32,33]. Consequently, there is need to independently validate the approach for specific regions and vegetation types [13,26,34,35] to determine if the technique is capable of inferring fire/burn severity from satellite imagery [31]. The Landsat NBR is used as a post-fire management tool in the USA and Canada, e.g., as operationally used by the Burned Area Emergency Rehabilitation (BAER) teams in the conterminous USA [12]. Numerous studies have demonstrated the usefulness of the index in the North American boreal and temperate regions [11,31,36,37,38], however, far fewer studies have assessed its effectiveness in California chaparral shrublands [9,20,39], an ecosystem which is highly sensitive to burning [39,40,41]. The few studies in the California chaparral shrublands demonstrated that the NBR is reasonably well related to fire severity, however, none of them conducted an inter-indices comparison. In Mediterranean shrublands in Europe similar findings were obtained by De Santis and Chuvieco [18], Veraverbeke et al. [32,33], Escuin et al. [42] and Tanase et al. [43]. Limited comparisons with other spectral indices were undertaken by De Santis and Chuvieco [18], Veraverbeke et al. [32,33] and Escuin et al. [42] concluding that the NBR outperformed the Normalized Difference Vegetation Index (NDVI) for assessing fire severity in Mediterranean shrublands. In addition, several authors indicated that the post-fire temperature increase as observed in the thermal infrared (TIR) domain [44,45] is complementary to the NBR for discriminating burned areas [46,47,48]. Holden et al. [46] demonstrated that enhancing the NBR with Landsat’s thermal band resulted in a better separability between burned and unburned land for two wildland fires in New Mexico, USA, whereas Veraverbeke et al. [48] revealed the potential of combining optical and thermal imagery for discriminating burned areas. In the study of Veraverbeke et al. [48] the prospect of enhancing the NBR with thermally derived surface emissivity was shown for separating burned areas in southern California. This study aims to evaluate the performance of existing vegetation indices and thermally enhanced indices for assessing fire severity in chaparral ecosystems. The study uses MODIS/ASTER (MASTER) airborne simulator data acquired over four southern California fire scars. MASTER was developed to support scientific studies by the Advanced Spaceborne Thermal Emission Radiometer (ASTER) and MODerate resolution Imaging Spectrodradiometer (MODIS) projects [49]. These high spatial and high spectral resolution data provide a unique opportunity for an in-depth evaluation of the sensitivities of several indices to fire severity in chaparral ecosystems.

2. Methodology

2.1. Study Area

Wildfires are a yearly recurring phenomena in southern California [20,41,50]. In the 2007 fire season 23 separately named fires burned over 200,000 ha. These fires mainly occurred in mountainous terrain and were driven by the strong Santa Ana winds, which are known to create extreme fire conditions in the region [51]. Four of the 2007 southern California wildfires were selected for this study. These fires are called the Canyon fire, the Harris fire, the Poomacha fire and the Witch fire (Figure 1). The fires occurred in October 2007 during a Santa Ana event. The main vegetation type consumed during these fires was chaparral. Chaparral is a complex and distinctive shrub formation that dominates the hills and lower mountain slopes of California. In southern California chaparral occupies approximately 70% of the land below the pine forests [50]. Composed of evergreen, woody shrubs, it often forms extensive and almost impenetrable stands. It is strongly adapted to protracted yearly drought and it is prone to recurrent fires, but vigorous in its recovery [20].

Figure 1. Location of the study sites in southern California.

Figure 1. Location of the study sites in southern California.

2.2. Field Data

Assessments of fire severity were obtained using the field protocol described in the Fire Monitoring Handbook (FMH, [52]). 25 field plots were established in November 2007, within one month after the fires. The plots were selected taking into account the constraints on mainly accessibility and time, encompassing the range of variability in fire severity in the study areas. To minimize potential misregistration the plots were GPS-recorded in relatively homogeneous areas in terms of post-fire effects. The FMH protocol is based on a visual rating assessment. Both fire effects on substrate and vegetation were separately judged. A description of the five fire severity classes for both substrate and vegetation is given in Table 1. Ratings range between five (high fire severity) and one (unburned). Figure 2 shows the distribution of the field plots over the different severity classes.

Figure 2. Distribution of the field plots over the fire severity (FS) classes.

Figure 2. Distribution of the field plots over the fire severity (FS) classes.

Table 1. Description of the fire severity classes for both substrate and vegetation (after [41]).

Fire severity ClassSubstrateVegetation
Unburned (1)Not burnedNot burned
Very low (2)Litter partially blackened; duff nearly unchanged; wood/leaf structures unchangedFoliage scorched and attached to supporting twigs
Low (3)Litter charred to partially consumed, some leaf structure undamaged; surface is predominantly black; some gray ash may be present immediately postburn; charring may extend slightly into soil surface where litter is sparse, otherwise soil is not alteredFoliage and smaller twigs partially to completely consumed; branches mostly intact; less than 60% of the shrub canopy is commonly consumed
Moderate (4)Leaf litter consumed, leaving coarse, light colored ash; duff deeply charred, but underlying mineral soil is not visibly altered; woody debris is mostly consumed; logs are deeply charred, burned-out stump holes are commonFoliage, twigs, ands small stems consumed; some branches (>0.6–1 cm in diameter) still present; 40–80% of the shrub canopy is commonly consumed
High (5)Leaf litter completely consumed, leaving a fluffy fine white ash; all organic material is consumed in mineral soil to a depth of 1–2.5 cm, this is underlain by a zone of black organic material; colloidal structure of the surface mineral soil may be alteredAll plant parts consumed leaving only stubs greater than 1 cm in diameter

2.3. MASTER Imagery and Preprocessing

The MASTER airborne simulator acquired high spectral and high spatial resolution imagery over the burned area in November 2007. The spatial resolution of MASTER data ranges from 5 m to 50 m depending on flying height. The pixel size of the data in this study varied between 6 and 8.5 m depending on the site. The instrument acquires radiance spectra between 0.4 μm and 13 μm in 50 spectral channels with 11 visible NIR (VNIR) channels, 14 SWIR channels and 25 mid infrared (MIR) and TIR channels.

The MASTER data were provided as level 1b geolocated calibrated radiance. Atmospheric and Topographic Correction for Airborne Imagery (ATCOR4) software was used to correct for the influence of the atmosphere, solar illumination and sensor viewing geometry [53]. ATCOR4 uses a large database containing the results of radiative transfer calculations based on MODTRAN4 code [53]. The standard ATCOR4 desert aerosol model was chosen. The visible through SWIR bands (1–25) were processed to surface reflectance, the MIR bands (26–40) were not atmospherically corrected and the thermal bands (41–50) were atmospherically corrected to surface radiance. The surface radiance of the thermal bands was then separated into surface temperature (Ts) and surface emissivity () using the emissivity normalization method [54]. is defined as the ratio of the actual emitted radiance to the radiance emitted from a blackbody at the same thermodynamic temperature [55,56]. Finally, the images were georeferenced using input geometry.

2.4. Spectral Indices

This study evaluates the performance of 19 spectral indices as listed in Table 2. This table includes: (i) the widely used NDVI [57], Global Environment Monitoring Index (GEMI, [58]) Enhanced Vegetation Index (EVI, [59]), Soil Adjusted Vegetation Index (SAVI, [60]) and Modified Soil Adjusted Vegetation Index (MSAVI, [61]); (ii) indices specifically designed for burned land applications such as the Burned Area Index (BAI, [62

Explore Doctor Jokes, Brain Fog, and more!

Haemangioblatomas are tumours of vascular origin and occur both in patients with von Hippel Lindau (vHL) as well as sporadically. They are WHO grade I tumours that can occur in the central nervous system or elsewhere in the body, including kidneys, liver and pancreas. Typically haemangioblastomas are sharply demarcated homogeneous masses composed of cyst with a solid enhancing mural nodule. http://radiopaedia.org/articles/haemangioblastoma-central-nervous-system

Medulloblastomas are a small round blue cell tumour and are the most common paediatric posterior fossa tumour and account for 30-40% of such entities. On non contrast CT, medulloblastomas appear as a mass : •arising from the vermis, •resulting in effacement of the fourth ventricle / basal cisterns and obstructive hydrocephalus. •usually hyperdense (90%) •cysts formation/necrosis is common (40-50%) •calcification in 10-20% of cases. http://radiopaedia.org/articles/medulloblastoma

One thought on “Lhermitte Evaluation Essay

Leave a comment

L'indirizzo email non verrà pubblicato. I campi obbligatori sono contrassegnati *