Search

Triangulating corpus linguistic approaches with other (linguistic and non-linguistic) approaches enhances “both the rigour of corpus linguistics and its incorporation into all kinds of research” (McEnery & Hardie, 2012: 227). Our study investigates an important area of mental health research: the experiences of those who hear voices that others cannot hear, and particularly the ways in which those voices are described as person-like. We apply corpus methods to augment the findings of a qualitative approach to 40 interviews with voice-hearers, whereby each interview was coded as involving ‘minimal’ or ‘complex’ personification of voices. Our analysis provides linguistic evidence in support of the qualitative coding of the interviews, but also goes beyond a binary approach by revealing different types and degrees of personification of voices, based on how they are referred to and described by voice-hearers. We relate these findings to concepts that inform therapeutic interventions in clinical psychology.

Background When fire intervals are shorter than the time required for plants to reproduce, plant populations are threatened by “immaturity risk.” Therefore, understanding how the time between fires influences plants can inform ecosystem management. Quantifying periods of immaturity risk requires investigating the influence of fire intervals across plant life stages, but most studies are indiscriminate of maturity. As fire regimes are multidimensional, it is also important to consider other characteristics of fires such as severity. We conducted a field study in heathy woodland that investigated how fire severity and fire interval influence immaturity risk to serotinous resprouter species, by examining if fire severity interacts with the time since the fire to influence the occurrence of mature individuals and relative abundance of three species: silver banksia (Banksia marginata Cav.), prickly teatree (Leptospermum continentale Joy Thomps), and heath teatree (Leptospermum myrsinoides Schitdl). Results Regression modeling revealed a strong, positive influence of time since the last fire on the proportion of quadrats at a site with mature plants, for all three species. We only detected a small and uncertain influence of fire severity on the proportion of quadrats with mature heath and prickly teatree, and did not observe an effect of fire severity on the maturity of silver banksia. Interestingly, no relationships were observed between time since fire and the relative abundance of plants. That is, only when plant life stages were considered did we detect an effect of fire on plants. Populations of the three species were mostly immature in the first 7 years post-fire, suggesting if sites were uniformly burnt in this time frame, there could be increased risk of local extinctions. Conclusions Our study highlights the importance of examining population processes, such as reproduction, in addition to plant relative abundance. Surprisingly, we did not detect strong differences in plant maturation across fire severity classes; low occurrence of mature plants in recently burnt areas indicated that immaturity risk was high, regardless of fire severity. Ecological studies that distinguish between plant life stages will help to predict the impacts of fire on populations and enhance decision-making. We recommend fire intervals of ≥ 8 years to protect serotinous resprouter plants in heathy woodland vegetation of southern Australia.

Record-breaking fire seasons in many regions across the globe raise important questions about plant community responses to shifting fire regimes (i.e., changing fire frequency, severity, and seasonality). Here, we examine the impacts of climate-driven shifts in fire regimes on vegetation communities, and likely responses to fire coinciding with severe drought, heatwaves and/or insect outbreaks. We present scenario-based conceptual models on how overlapping disturbance events and shifting fire regimes interact differently to limit post-fire resprouting and recruitment capacity. We demonstrate that, although many communities will remain resilient to changing fire regimes in the short-term, longer-term changes to vegetation structure, demography and species composition are likely, with a range of subsequent effects on ecosystem function. Resprouting species are likely to be most resilient to changing fire regimes. However, even these species are susceptible if exposed to repeated short-interval fire in combination with other stressors. Post-fire recruitment is highly vulnerable to increased fire frequency, particularly as climatic limitations on propagule availability intensify. Prediction of community responses to fire under climate change will be greatly improved by addressing knowledge gaps on how overlapping disturbances and climate change-induced shifts in fire regime affect post-fire resprouting, recruitment, growth rates, and species-level adaptation capacity. This article is protected by copyright. All rights reserved.

This chapter takes a linguistic approach to the Voices in Psychosis interviews. It shows how a computer-aided linguistic analysis of the complete set of interviews reveals an unusually high frequency of negation. The chapter then focuses specifically on references to speech not taking place, and identifies four main types of descriptions of ‘silences’ in the interviews: silent ‘voices’; the voice-hearer being silent about the voices; the voice-hearer being silenced by the voices; and the voice-hearer being silent with the voices. The chapter links these four types of silences to broader issues to do with social relationships, control, and the phenomenology of voice-hearing.

In this paper, we reflect on the process of re-operationalising transcript data generated in an ethnographic study for the purposes of corpus analysis. We present a corpus of patient–provider interactions in the context of Emergency Departments in hospitals in Australia, to discuss the process through which ethnographic transcripts were manipulated to generate a searchable corpus. We refer to the types of corpus analysis that this conversion enables, facilitated by the rich metadata collected alongside the transcribed audio recordings, augmenting the findings of prior qualitative analyses. Subsequently, we offer guidance for spoken data transcription, intended to ‘future proof’ such data for subsequent reformatting for corpus linguistic analysis.

Elevated atmospheric [CO2] (‘eCO2’) may alter species composition within vegetation types by favouring the growth of some species over others. However, other related changes in climate conditions, such as increased frequency and severity of drought, may reduce eCO2 fertilisation effects on plant growth. For many species, it is not known if responses will reflect variability in trait adaptations due to environment. We grew seedlings of nine species of eucalypts indicative of three regional vegetation types (representing a mesic–xeric ecosystem gradient) under two CO2 concentrations (400 parts per million; 640 ppm, i.e. eCO2) and two watering regimes (well-watered; drought-like conditions). Elevated CO2 increased biomass accumulation but drought reduced this effect, with mesic species experiencing larger relative reductions. Elevated CO2 increased the size of storage organs used during resprouting, in the absence of drought. Typical drought responses, such as increased leaf mass per unit area and root mass ratio, were more pronounced in xeric species and were reduced under eCO2. Seedling growth and resprouting may be enhanced by eCO2, suggesting continued dominance of resprouting species in disturbance-prone ecosystems, although severe drought is likely to offset eCO2 fertilisation. Xeric species may cope with drought more effectively under eCO2 than mesic species due to resource acquisition and storage traits that are more responsive. Read the free Plain Language Summary for this article on the Journal blog.

Elevated tree mortality and reduced recruitment of new trees linked to drought and fires has been reported across a range of forests over the last few decades. Forests that resprout new foliage epicormically from buds beneath the bark are considered highly resilient to disturbance, but are potentially at risk of elevated mortality, demographic shifts and changes to species composition due to synergistic effects of drought and fire. Despite this, the effects of drought-fire interactions on such forests remain largely unknown. We assessed the effects of drought severity and fire frequency on juvenile mortality, post-fire seedling recruitment and replacement of juvenile trees (balance of recruitment minus mortality) following fire. We compared dry ridgetop and wet gully assemblages across a temperate forest in southern Australia. Both forest types experienced higher rates of fire-induced juvenile mortality in areas that had experienced severe drought compared to moderate drought, though mortality rates were generally low across all drought and fire combinations. This result indicated that topographic position (i.e. wet gullies) did little to moderate juvenile mortality when exposed to severe drought plus fire. In wet forest, severe drought also reduced recruitment and replacement of dead juveniles by post-fire seedlings compared to moderate drought. In dry forest net-negative replacement increased with the severity of drought. Across both forest types, the total pool of juveniles was reduced under severe drought. Future increases in the frequency of coupled severe drought and fire will likely increase the susceptibility of resilient temperate forests to major changes in structure and function.

Research highlights—Feedbacks between fire severity, vegetation structure and ecosystem flammability are understudied in highly fire-tolerant forests that are dominated by epicormic resprouters. We examined the relationships between the severity of two overlapping fires in a resprouting eucalypt forest and the subsequent effect of fire severity on fuel structure. We found that the likelihood of a canopy fire was the highest in areas that had previously been exposed to a high level of canopy scorch or consumption. Fuel structure was sensitive to the time since the previous canopy fire, but not the number of canopy fires. Background and Objectives—Feedbacks between fire and vegetation may constrain or amplify the effect of climate change on future wildfire behaviour. Such feedbacks have been poorly studied in forests dominated by highly fire-tolerant epicormic resprouters. Here, we conducted a case study based on two overlapping fires within a eucalypt forest that was dominated by epicormic resprouters to examine (1) whether past wildfire severity affects future wildfire severity, and (2) how combinations of understorey fire and canopy fire within reburnt areas affect fuel properties. Materials and Methods—The study focused on ≈77,000 ha of forest in south-eastern Australia that was burnt by a wildfire in 2007 and reburnt in 2013. The study system was dominated by eucalyptus trees that can resprout epicormically following fires that substantially scorch or consume foliage in the canopy layer. We used satellite-derived mapping to assess whether the severity of the 2013 fire was affected by the severity of the 2007 fire. Five levels of fire severity were considered (lowest to highest): unburnt, low canopy scorch, moderate canopy scorch, high canopy scorch and canopy consumption. Field surveys were then used to assess whether combinations of understorey fire (&lt, 80% canopy scorch) and canopy fire (&gt, 90% canopy consumption) recorded over the 2007 and 2013 fires caused differences in fuel structure. Results—Reburn severity was influenced by antecedent fire severity under severe fire weather, with the likelihood of canopy-consuming fire increasing with increasing antecedent fire severity up to those classes causing a high degree of canopy disturbance (i.e., high canopy scorch or canopy consumption). The increased occurrence of canopy-consuming fire largely came at the expense of the moderate and high canopy scorch classes, suggesting that there was a shift from crown scorch to crown consumption. Antecedent fire severity had little effect on the severity patterns of the 2013 fire under nonsevere fire weather. Areas affected by canopy fire in 2007 and/or 2013 had greater vertical connectivity of fuels than sites that were reburnt by understorey fires, though we found no evidence that repeated canopy fires were having compounding effects on fuel structure. Conclusions—Our case study suggests that exposure to canopy-defoliating fires has the potential to increase the severity of subsequent fires in resprouting eucalypt forests in the short term. We propose that the increased vertical connectivity of fuels caused by resprouting and seedling recruitment were responsible for the elevated fire severity. The effect of antecedent fire severity on reburn severity will likely be constrained by a range of factors, such as fire weather.

The main eigenvalues of a graph G are those eigenvalues of the (0, 1)-adjacency matrix A with a corresponding eigenspace not orthogonal to j = (1 | 1 | · · · | 1)T. The principal main eigenvector associated with a main eigenvalue is the orthogonal projection of the corresponding eigenspace onto j. The main eigenspace of a graph is generated by all the principal main eigenvectors and is the same as the image of the walk matrix. We explore a new concept to see to what extent the main eigenspace determines the entries of the walk matrix of a graph. The CDC of a graph G is the direct product G × K2. We establish a hierarchy of inclusions connecting classes of graphs in view of their CDC, walk matrix, main eigenvalues and main eigenspaces. We provide a new proof that graphs with the same CDC are characterized as TF-isomorphic graphs. A complete list of TF-isomorphic graphs on at most 8 vertices and their common CDC is also given., peer-reviewed

Reliable fire severity mapping is a vital resource for fire scientists and land management agencies globally. Satellite derived pre- and post-fire differenced severity indices (∆FSI), such as the differenced Normalised Burn Ratio (∆NBR), are widely used to map the severity of large wildfires. Fire severity classification is commonly undertaken through the identification of severity class thresholds in ∆FSI. However, several shortcomings have been identified with severity classifications using ∆FSI, including poor delineation of low fire severity classes, and unsatisfactory performance when ∆FSI classification thresholds are applied to new landscapes. Our study assesses the performance of the Random Forest classifier (RF) for improving the accuracy of satellite based wildfire severity mapping across heterogeneous landscapes using Landsat imagery. We collected point based fire severity training data (n = 10,855) from sixteen large wildfires occurring across south-eastern Australia between 2006 and 2016. The predictive accuracy of fire severity classification using ∆NBR and the RF incorporating numerous spectral indices, was assessed using bootstrapping and cross validation. Image acquisition and index calculation for each fire was undertaken in Google Earth Engine (GEE). Results of the bootstrapping validation show that the RF classifier had very high classification accuracy (>95%) for unburnt (UB), crown scorch (CS) and crown consumption (CC) severity classes, and high classification accuracy (>74%) for low severity classes (crown unburnt, CU; partial crown scorch, PCS). The RF classification outperformed the ∆NBR classification for all severity classes, increasing classification accuracy by between 6%–21%. Cross validation using independent fires produced similar median classification accuracy as the bootstrapping validation, though the RF classification of CU was substantially reduced to ~55%. ∆NBR, ∆NDWI and ∆NDVI were the three most important variables in the RF model. The Landsat satellite platform used to derive the indices had little effect on classification accuracy. Maps produced using the RF classifier in GEE had similar spatial patterns in fire severity classes as maps produced using time-consuming hand digitisation of aerial images. GEE was found to be a highly efficient platform for image acquisition, processing and production of severity maps. Our study shows that fire severity mapping using RF classifiers provides a robust method for broad scale mapping of fire severity across heterogeneous landscapes. Furthermore, the GEE-based classification framework supports the operational application of this approach in a land management agency context for the rapid production of fire severity maps.

We investigate the time-varying correlation between oil prices and stock prices. Estimation results from a multivariate DCC-GARCH model reveal that the correlation has changed since the financial c...

The storage of carbon in plant tissues and debris has been proposed as a method to offset anthropogenic increases in atmospheric [CO2 ]. Temperate forests represent significant above-ground carbon (AGC) "sinks" because their relatively fast growth and slow decay rates optimise carbon assimilation. Fire is a common disturbance event in temperate forests globally that should strongly influence AGC because: discrete fires consume above-ground biomass releasing carbon to the atmosphere, and the long-term application of different fire-regimes select for specific plant communities that sequester carbon at different rates. We investigated the latter process by quantifying AGC storage at 104 sites in the Sydney Basin Bioregion, Australia, relative to differences in components of the fire regime: frequency, severity and interfire interval. To predict the potential impacts of future climate change on fire/AGC interactions, we stratified our field sites across gradients of mean annual temperature and precipitation and quantified within- and between-factor interactions between the fire and climate variables. In agreement with previous studies, large trees were the primary AGC sink, accounting for ~70% of carbon at sites. Generalised additive models showed that mean annual temperature was the strongest predictor of AGC storage, with a 54% near-linear decrease predicted across the 6.1°C temperature range experienced at sites. Mean annual precipitation, fire frequency, fire severity and interfire interval were consistently poor predictors of total above-ground storage, although there were some significant relationships with component stocks. Our results show resilience of AGC to frequent and severe wildfire and suggest temperature mediated decreases in forest carbon storage under future climate change predictions.

A trend of increasing woody plant density, or woody thickening, has been observed across grassland and woodland ecosystems globally. It has been proposed that increasing atmospheric [CO2] is a major driver of broad scale woody thickening, though few field-based experiments have tested this hypothesis. Our study utilises a Free Air CO2 Enrichment experiment to examine the effect of elevated [CO2] (eCO2) on three mechanisms that can cause woody thickening, namely (i) woody plant recruitment, (ii) seedling growth, and (iii) post-disturbance resprouting. The study took place in a eucalypt-dominated temperate grassy woodland. Annual assessments show that juvenile woody plant recruitment occurred over the first 3 years of CO2 fumigation, though eCO2 did not affect rates of recruitment. Manipulative experiments were established to examine the effect of eCO2 on above-ground seedling growth using transplanted Eucalyptus tereticornis (Myrtaceae) and Hakea sericea (Proteaceae) seedlings. There was no positive effect of eCO2 on biomass of either species following 12 months of exposure to treatments. Lignotubers (i.e., resprouting organs) of harvested E. tereticornis seedlings that were retained in situ for an additional year were used to examine resprouting response. The likelihood of resprouting and biomass of resprouts increased with lignotuber volume, which was not itself affected by eCO2. The presence of herbaceous competitors and defoliation by invertebrates and pathogens were found to greatly reduce growth and/or resprouting response of seedlings. Our findings do not support the hypothesis that future increases in atmospheric [CO2] will, by itself, promote woody plant recruitment in eucalypt-dominated temperate grassy woodlands.

WHAT'S WRONG WITH women's nipples? Or rather, why do some social-media platforms allow men to post topless pictures of themselves, but ban women from doing the same thing? The answer to this question neatly encapsulates the key issue of online censorship today. We may think of the online world as a global commons, a public space for free expression, but it is increasingly a series of private spaces within which we are given limited permission to act. Our access to each other and to the resources available within these spaces is increasingly being controlled not just by law, but by gatekeepers whose motivations may include political advantage, social control, or simple profit. For example, the Free The Nipple movement was launched in 2012 to highlight the differing ways in which society expects the bodies of women and men to be portrayed, as part of a campaign focused on "equality, empowerment and the freedom of all human beings". When women started posting topless images of themselves using the #freethenipple hashtag on Instagram, their posts were banned for breaching the service's Community Guidelines. In a 2015 interview with Business Insider, Kevin Systrom, CEO of Instagram, explained that one reason for this was to meet the content guidelines of Apple's App Store. One gatekeeper (Instagram) bowed to another (Apple) to control how a campaign for equality was expressed online. In a sense, this should be no surprise. When Steve Jobs launched the iPhone App Store in 2008, he said that apps would not be allowed to distribute pornography, hog bandwidth, breach user privacy, act maliciously, or otherwise violate rules set by the company. The motivation, Jobs said, was "to get a ton of apps out there". This is fair enough in the context of furthering a business' interests. However, as the web, Facebook, YouTube, Instagram, Twitter, WhatsApp, Sina Weibo, Snapchat and more gather billions of users and become prime conduits for our communications, the way they are governed and controlled becomes increasingly important to the functioning of society. How big an issue is online censorship? According to estimates by Freedom House, a US watchdog backed by sponsors ranging from the US State Department and the Dutch Ministry of Foreign Affairs to BAE Systems and Google: . 3.2 billion people have access to the internet . 67 per cent live in countries where criticism of the government, military, or ruling family has been subject to censorship . 60 per cent live in countries where ICT users were arrested or imprisoned for posting content on political, social and religious issues . 49 per cent live in countries where individuals have been attacked or killed for their online activities since June 2015 . 47 per cent live in countries where insulting religion online can result in censorship or jail time . 33 per cent live in countries where online discussion of LGBT+ issues can be repressed or punished . 38 per cent live in countries where social media or messaging apps were blocked over the past year . 27 per cent live in countries THIS IIIS MRCI'IIME FIGHTS IMTERλeET SHUTDOWMS #Keej On The pressure group Access Now monitors and campaigns against internet shutdowns Eπgiπaartnthetas 8 Tichnology November 2017 www.EandTmagazine ceπwhere users have been arrested for writing, sharing or even liking Facebook posts . 38 per cent live under governments that disconnected internet or mobile phone access, often for political reasons. In other words, the global commons is increasingly strictly policed, and offending its gatekeepers can lead to harsh punishments.

Atmospheric carbon dioxide enrichment (eCO2) can enhance plant carbon uptake and growth1–5, thereby providing an important negative feedback to climate change by slowing the rate of increase of the atmospheric CO2 concentration6. Although evidence gathered from young aggrading forests has generally indicated a strong CO2 fertilization effect on biomass growth3–5, it is unclear whether mature forests respond to eCO2 in a similar way. In mature trees and forest stands7–10, photosynthetic uptake has been found to increase under eCO2 without any apparent accompanying growth response, leaving the fate of additional carbon fixed under eCO2 unclear4,5,7–11. Here using data from the first ecosystem-scale Free-Air CO2 Enrichment (FACE) experiment in a mature forest, we constructed a comprehensive ecosystem carbon budget to track the fate of carbon as the forest responded to four years of eCO2 exposure. We show that, although the eCO2 treatment of +150 parts per million (+38 per cent) above ambient levels induced a 12 per cent (+247 grams of carbon per square metre per year) increase in carbon uptake through gross primary production, this additional carbon uptake did not lead to increased carbon sequestration at the ecosystem level. Instead, the majority of the extra carbon was emitted back into the atmosphere via several respiratory fluxes, with increased soil respiration alone accounting for half of the total uptake surplus. Our results call into question the predominant thinking that the capacity of forests to act as carbon sinks will be generally enhanced under eCO2, and challenge the efficacy of climate mitigation strategies that rely on ubiquitous CO2 fertilization as a driver of increased carbon sinks in global forests. Carbon dioxide enrichment of a mature forest resulted in the emission of the excess carbon back into the atmosphere via enhanced ecosystem respiration, suggesting that mature forests may be limited in their capacity to mitigate climate change.

Atmospheric carbon dioxide enrichment (eCO2) can enhance plant carbon uptake and growth1,2,3,4,5, thereby providing an important negative feedback to climate change by slowing the rate of increase of the atmospheric CO2concentration6. While evidence gathered from young aggrading forests has generally indicated a strong CO2fertilization effect on biomass growth3,4,5, it is unclear whether mature forests respond to eCO2in a similar way. In mature trees and forest stands7,8,9,10, photosynthetic uptake has been found to increase under eCO2without any apparent accompanying growth response, leaving an open question about the fate of additional carbon fixed under eCO24, 5, 7,8,9,10,11. Here, using data from the first ecosystem-scale Free-Air CO2Enrichment (FACE) experiment in a mature forest, we constructed a comprehensive ecosystem carbon budget to track the fate of carbon as the forest responds to four years of eCO2exposure. We show that, although the eCO2treatment of ambient +150 ppm (+38%) induced a 12% (+247 gCm-2yr-1) increase in carbon uptake through gross primary production, this additional carbon uptake did not lead to increased carbon sequestration at the ecosystem level. Instead, the majority of the extra carbon was emitted back into the atmosphere via several respiratory fluxes, with increased soil respiration alone contributing ∼50% of the total uptake surplus. Our results call into question the predominant thinking that the capacity of forests to act as carbon sinks will be generally enhanced under eCO2, and challenge the efficacy of climate mitigation strategies that rely on CO2fertilization as a driver of increased carbon sinks in standing forests and afforestation projects.

Satellite-derived spectral indices such as the relativized burn ratio (RBR) allow fire severity maps to be produced in a relatively straightforward manner across multiple fires and broad spatial extents. These indices often have strong relationships with field-based measurements of fire severity, thereby justifying their widespread use in management and science. However, satellite-derived spectral indices have been criticized because their non-standardized units render them difficult to interpret relative to on-the-ground fire effects. In this study, we built a Random Forest model describing a field-based measure of fire severity, the composite burn index (CBI), as a function of multiple spectral indices, a variable representing spatial variability in climate, and latitude. CBI data primarily representing forested vegetation from 263 fires (8075 plots) across the United States and Canada were used to build the model. Overall, the model performed well, with a cross-validated R2 of 0.72, though there was spatial variability in model performance. The model we produced allows for the direct mapping of CBI, which is more interpretable compared to spectral indices. Moreover, because the model and all spectral explanatory variables were produced in Google Earth Engine, predicting and mapping of CBI can realistically be undertaken on hundreds to thousands of fires. We provide all necessary code to execute the model and produce maps of CBI in Earth Engine. This study and its products will be extremely useful to managers and scientists in North America who wish to map fire effects over large landscapes or regions.

Ecosystem management can negatively affect the demography of plant communities through the introduction of novel disturbance regimes. Prescribed burning and timber harvesting are two common and widely applied management strategies across forest ecosystems. Despite this, little is known about the long-term effects that these interacting disturbances have on forest demography. This study examined the effect of timber harvesting and frequent prescribed burning on the mortality, growth and regeneration of trees in a temperate eucalypt forest of south-eastern Australia. The study took place at a long-term experimental site, where experimental coupes were subjected to a one-off selective harvesting treatment (harvested, not harvested), followed by regimes of experimental burning (no fire, ~4 year burn intervals or ~2 year burn intervals) over a 22 year period. Tree communities were surveyed at permanent monitoring sites prior to the application of experimental treatments (1985 – 1989), and resurveyed post treatment (2016) to assess mortality, growth rates and ingrowth of trees >10 cm diameter at breast height. Harvesting directly removed ~40% of trees and indirectly increased the mortality of retained trees through damage (e.g. crown and bole breakage) caused during the harvesting operation. The likelihood of harvesting damage was greater for small trees and increased with harvesting intensity (i.e. the amount of timber removed). Frequent burning increased the likelihood of tree mortality on harvested sites, with large, old trees being particularly vulnerable. Growth rate and ingrowth of trees was elevated at harvested sites, increasing almost linearly with harvesting intensity, which suggests that competitive release had occurred. Fire frequency had no effect on growth rates or ingrowth of trees. This study highlights that frequent prescribed burning and selective timber harvesting can have additive effects on the loss of large trees, reducing the availability of these keystone habitat structures in intensively managed forest ecosystems. Although the elevated rates of growth and ingrowth may hasten the replacement of lost large trees, recovery will require long time frames.

The ability to map fire severity is a requirement for fire management agencies worldwide. The development of repeatable methods to produce accurate and consistent fire severity maps from satellite imagery is necessary to document fire regimes, to set priorities for post-fire management responses, and for research applications. Machine learning techniques, such as random forest, have shown great promise for mapping of wildfire severity in woodland and forest ecosystems using satellite imagery. However, an assessment of the properties of training data required for automated mapping with random forest is currently lacking. This study examined how training data properties affect fire severity classification across forest, woodland and shrubland communities of southern Australia. The aims of this study were: (i) to examine how sample size (i.e. number of training points and fire events) and sample imbalance affect classification accuracy; (ii) to determine whether models were transferrable across geographic regions; and (iii) to assess the need for classifiers for prescribed burns and wildfires. We sampled 33 wildfires and 57 prescribed burns occurring across southern Australia between 2006 and 2019, to derive an extensive dataset (n = 25, 350 points) suitable for model training and validation. Five fire severity classes were mapped across the forest, woodland and shrubland communities. Using independent spatial cross validation of wildfires, we found that a minimum of 300 sample points per severity class, sampled across at least ten independent fires, was sufficient to reach the upper threshold of classification accuracy for the five severity classes. Training datasets derived across broad environmental space or close (≤100 km) to the target fire (i.e. fire to be mapped) produced better predictions than those derived regionally and far (>100 km) from the target fire. Models trained on data derived from both wildfires and prescribed burns had similar overall accuracy as those trained only on data from the fire type being predicted. However, there were significant differences in accuracy between the models trained with wildfires, prescribed burns and combined datasets within some severity classes. The random forest classifier had an overall accuracy of ~88% for wildfires and ~68% for prescribed burns across the study fires. The discrepancy in accuracy between wildfires and burns was likely due to the poorer classification performance for low fire severity classes, the dominant severity classes in prescribed burns. Overall classification accuracy was relatively consistent across forest, woodland and shrubland communities within the study fires, indicating that the method is robust across the temperate forest biome of southern Australia. Our results demonstrate that consideration of training data properties, specifically the number of points and fires sampled, sample balance and the geographic source of sample data, will be important considerations for the automated mapping of fire severity using random forest classification and Landsat imagery.

The ISUP (Internal Society of Urologic Pathology) recently adopted a five-tiered prognostication system. There is evidence to suggest that the ISUP grade group 4 is a heterogeneous entity regarding prognosis. Our aim was to systematically examine the existing evidence to determine if outcome differences exist within the ISUP grade group 4. A systematic search of the literature for all studies examining the heterogeneity of the ISUP grade group 4 was conducted. Available studies were combined with meta-analysis to evaluate prognostic differences within the ISUP grade group 4 measured by all-cause mortality (ACM) and the prostate cancer-specific mortality (PCSM). Eight studies were identified and utilised a variety of outcome measures to answer the question of heterogeneity within the ISUP grade group 4. Four of these studies examined prognosis using both ACM and PCSM. These were combined into a meta-analysis. The combined group of 5 + 3/3 + 5 had statistically significant higher ACM (hazard ratio [HR] 1.23, 95% confidence internal [Cl] 1.08–1.41) when compared to the 4 + 4 group. There was no difference in the PCSM between the two groups (HR 1.34, 95% CI 0.89–2.01). However, heterogeneity was high for this analysis secondary to a range of methodological differences. Our meta-analysis showed that Gleason grade 3 + 5/5 + 3 had higher ACM than Gleason grade group 4 + 4. Measures of PCSM were statistically insignificant, although heterogeneity was high. Evidence suggests that heterogeneity is likely, although inconclusive. Further studies with consistent methodologies are required to answer this question.

Wildfire refugia (unburnt patches within large wildfires) are important for the persistence of fire-sensitive species across forested landscapes globally. A key challenge is to identify the factors that determine the distribution of fire refugia across space and time. In particular, determining the relative influence of climatic and landscape factors is important in order to understand likely changes in the distribution of wildfire refugia under future climates. Here, we examine the relative effect of weather (i.e. fire weather, drought severity) and landscape features (i.e. topography, fuel age, vegetation type) on the occurrence of fire refugia across 26 large wildfires in south-eastern Australia. Fire weather and drought severity were the primary drivers of the occurrence of fire refugia, moderating the effect of landscape attributes. Unburnt patches rarely occurred under 'severe' fire weather, irrespective of drought severity, topography, fuels or vegetation community. The influence of drought severity and landscape factors played out most strongly under 'moderate' fire weather. In mesic forests, fire refugia were linked to variables that affect fuel moisture, whereby the occurrence of unburnt patches decreased with increasing drought conditions and were associated with more mesic topographic locations (i.e. gullies, pole-facing aspects) and vegetation communities (i.e. closed-forest). In dry forest, the occurrence of refugia was responsive to fuel age, being associated with recently burnt areas (

Accurate and consistent broad-scale mapping of fire severity is an important resource for fire management as well as fire-related ecological and climate change research. Remote sensing and machine learning approaches present an opportunity to enhance accuracy and efficiency of current practices. Quantitative biophysical models of photosynthetic, non-photosynthetic and bare cover fractions have not been widely applied to fire severity studies but may provide greater consistency in comparisons of different fires across the landscape compared to reflectance-based indices. We systematically tested and compared reflectance and fractional cover candidate severity indices derived from Sentinel 2 satellite imagery using a random forest (RF) machine learning framework. Assessment of predictive power (cross-validation) was undertaken to quantify the accuracy of mapping severity of new fires. The effect of environmental variables on the accuracy of the RF predicted severity classification was examined to assess the stability of the mapping across the landscape. The results indicate that fire severity can be mapped with very high accuracy using Sentinel 2 imagery and RF supervised classification. The mean accuracy was >95% for the unburnt and extreme severity class (complete crown consumption), >85% for high severity class (full crown scorch), >80% for low severity (burnt understory, unburnt canopy) and >70% for the moderate severity class (partial canopy scorch). Higher canopy cover and higher topographic complexity was associated with a higher rate of under-prediction, due to the limitations of optical sensors in viewing the burnt understorey of low severity classes under these conditions. Further research is aimed at improving classification accuracy of low and moderate severity classes and applying the RF algorithm to hazard reduction fires.