Geographic Analysis of Forest Health Indicators Using Spatial Scan Statistics

Geographically explicit analysis tools are needed to assess forest health indicators that are measured over large regions. Spatial scan statistics can be used to detect spatial or spatiotemporal clusters of forests representing hotspots of extreme indicator values. This paper demonstrates the approach through analyses of forest fragmentation indicators in the southeastern United States and insect and pathogen indicators in the Pacific Northwest United States. The scan statistic detected four spatial clusters of fragmented forest including a hotspot in the Piedmont and Coastal Plain region. Three recurring clusters of insect and pathogen occurrence were found in the Pacific Northwest. Spatial scan statistics are a powerful new tool that can be used to identify potential forest health problems.     

Using Satellite Fire Detection to Calibrate Components of the Fire Weather Index System in Malaysia and Indonesia

Vegetation fires have become an increasing problem in tropical environments as a consequence of socioeconomic pressures and subsequent land-use change. In response, fire management systems are being developed. This study set out to determine the relationships between two aspects of the fire problems in western Indonesia and Malaysia, and two components of the Canadian Forest Fire Weather Index System. The study resulted in a new method for calibrating components of fire danger rating systems based on satellite fire detection (hotspot) data. Once the climate was accounted for, a problematic number of fires were related to high levels of the Fine Fuel Moisture Code. The relationship between climate, Fine Fuel Moisture Code, and hotspot occurrence was used to calibrate Fire Occurrence Potential classes where low accounted for 3% of the fires from 1994 to 2000, moderate accounted for 25%, high 26%, and extreme 38%. Further problems arise when there are large clusters of fires burning that may consume valuable land or produce local smoke pollution. Once the climate was taken into account, the hotspot load (number and size of clusters of hotspots) was related to the Fire Weather Index. The relationship between climate, Fire Weather Index, and hotspot load was used to calibrate Fire Load Potential classes. Low Fire Load Potential conditions (75% of an average year) corresponded with 24% of the hotspot clusters, which had an average size of 30% of the largest cluster. In contrast, extreme Fire Load Potential conditions (1% of an average year) corresponded with 30% of the hotspot clusters, which had an average size of 58% of the maximum. Both Fire Occurrence Potential and Fire Load Potential calibrations were successfully validated with data from 2001. This study showed that when ground measurements are not available, fire statistics derived from satellite fire detection archives can be reliably used for calibration. More importantly, as a result of this work, Malaysia and Indonesia have two new sources of information to initiate fire prevention and suppression activities.     

Biophysical and anthropogenic controls of forest fires in the Deccan Plateau, India

Forest fires constitute one of the most serious environmental problems in several forested regions of India. In the Indian sub-continent, relatively few studies have focused on the assessment of biophysical and anthropogenic controls of forest fires at a landscape scale and the spatial aspects of these relationships. In this study, we used fire count data sets from satellite remote sensing data covering 78 districts over four different states of the Deccan Plateau, India, for assessing the underlying causes of fires. Spatial data for explanatory variables of fires pertaining to topography, vegetation, climate, anthropogenic and accessibility factors have been gathered corresponding with fire presence/absence. A logistic regression model was used to estimate the probability of the presence of fires as a function of the explanatory variables. Results for fire area estimates suggested that, of the total fires covering 47,043km(2) that occurred during the year 2000 for the entire Indian region, 29.0% occurred in the Deccan Plateau, with Andhra Pradesh having 13.5%, Karnataka 14.7%, Kerala 0.1%, and Tamilnadu 1.15%. Results from the logistic regression suggest that the strongest influences on the fire occurrences were the amount of forest area, biomass densities, rural population density (PD), average precipitation of the warmest quarter, elevation (ELE) and mean annual temperature (MAT). Among these variables, biomass density (BD) and average precipitation of the warmest quarter had the highest significance, followed by others. These results on the best predictors of forest fires can be used both as a strategic planning tool to address broad scale fire risk concerns, and also as a tactical guide to help forest managers to design fire mitigation measures at the district level.     

Locating Spatial Variation in the Association Between Wildland Fire Risk and Social Vulnerability Across Six Southern States

Wildland fire in the South commands considerable attention, given the expanding wildland urban interface (WUI) across the region. Much of this growth is propelled by higher income retirees and others desiring natural amenity residential settings. However, population growth in the WUI increases the likelihood of wildfire fire ignition caused by people, as humans account for 93% of all wildfires fires in the South. Coexisting with newly arrived, affluent WUI populations are working class, poor or otherwise socially vulnerable populations. The latter groups typically experience greater losses from environmental disasters such as wildfire because lower income residents are less likely to have established mitigation programs in place to help absorb loss. We use geographically weighted regression to examine spatial variation in the association between social vulnerability (SOVUL) and wildfire risk. In doing so, we identify “hot spots” or geographical clusters where SOVUL varies positively with wildfire risk across six Southern states—Alabama, Arkansas, Florida, Georgia, Mississippi, and South Carolina. These clusters may or may not be located in the WUI. These hot spots are most prevalent in South Carolina and Florida. Identification of these population clusters can aid wildfire managers in deciding which communities to prioritize for mitigation programming.     

A Review of the Main Driving Factors of Forest Fire Ignition Over Europe

Knowledge of the causes of forest fires, and of the main driving factors of ignition, is an indispensable step towards effective fire prevention policies. This study analyses the factors driving forest fire ignition in the Mediterranean region including the most common human and environmental factors used for modelling in the European context. Fire ignition factors are compared to spatial and temporal variations of fire occurrence in the region, then are compared to results obtained in other areas of the world, with a special focus on North America (US and Canada) where a significant number of studies has been carried out on this topic. The causes of forest fires are varied and their distribution differs among countries, but may also differ spatially and temporally within the same country. In Europe, and especially in the Mediterranean basin, fires are mostly human-caused mainly due arson. The distance to transport networks and the distance to urban or recreation areas are among the most frequently used human factors in modelling exercises and the Wildland-Urban Interface is increasingly taken into account in the modelling of fire occurrence. Depending on the socio-economic context of the region concerned, factors such as the unemployment rate or variables linked to agricultural activity can explain the ignition of intentional and unintentional fires. Regarding environmental factors, those related to weather, fuel and topography are the most significant drivers of ignition of forest fires, especially in Mediterranean-type regions. For both human and lightning-caused fires, there is a geographical gradient of fire ignition, mainly due to variations in climate and fuel composition but also to population density for instance. The timing of fires depends on their causes. In populated areas, the timing of human-caused fires is closely linked to human activities and peaks in the afternoon whereas, in remote areas, the timing of lightning-caused fires is more linked to weather conditions and the season, with most such fires occurring in summer.     

Remotely-Sensed Active Fire Data for Protected Area Management: Eight-Year Patterns in the Manas National Park, India

The Terai-Duar savanna and grasslands, which once extended along most of the Himalayan foothills, now only remain in a number of protected areas. Within these localities, grassland burning is a major issue, but data on frequency and distribution of fires are limited. Here, we analysed the incidence of active fires, which only occur during the dry season (Nov.–Mar.), within a significant area of Terai grasslands: the Manas National Park (MNP), India. We obtained locations of 781 fires during the 2000–2008 dry seasons, from the Fire Information for Resource Management System (FIRMS) that delivers global MODIS hotspot/fire locations using remote sensing and GIS technologies. Annual number of fires rose significantly from around 20 at the start of the study period to over 90 after 2002, with most (85%) detected between December and January. Over half of the fires occurred in tall grasslands, but fire density was highest in wetland and riverine vegetation, dry at the time. Most burning took place near rivers, roads and the park boundary, suggesting anthropogenic origins. A kernel density map of all recorded fires indicated three heavily burnt areas in the MNP, all within the tall grasslands. Our study demonstrates, despite some technical caveats linked to fire detection technology, which is improving, that remote fire data can be a practical tool in understanding fire concentration and burning temporal patterns in highly vulnerable habitats, useful in guiding management.     

Predicting Sustained Fire Spread in Tasmanian Native Grasslands

Fire is widely used in conservation management of native grasslands. Burning is often carried out under conditions that are marginal for sustained fire spread, and therefore it would be useful to be able to predict fire sustainability. There is currently no model allowing such prediction in temperate grasslands. This study aims to identify the environmental variables that determine whether fires will sustain in native grasslands in Tasmania, Australia, and develop a model for predicting fire sustainability in this vegetation. Fuel characteristics and weather conditions were recorded for 111 test fires. Logistic regression modeling identified dead fuel moisture content, fuel load, and percentage dead fuel as predictors of fire sustainability. Classification tree modeling identified dead fuel moisture and fuel load threshold values for sustaining fires. There was also evidence indicating a percentage dead fuel threshold. The logistic regression model and a model combining the results of the classification tree and the percentage dead fuel threshold accurately predicted the outcomes of a small set of experimental fires. These models are likely to have utility in predicting fire sustainability in Tasmanian grasslands and are also likely to be applicable to similar grasslands elsewhere.     

CRITICAL DATA REQUIREMENTS FOR PREDICTION OF EROSION AND SEDIMENTATION IN MOUNTAIN DRAINAGE BASINS1

ABSTRACT: The potential for understanding and, where necessary, managing sedimentation in humid mountain drainage basins increases with awareness of the conditions that lead to shallow landsliding, debris flows, and catastrophic sedimentation in stream channels. Progress in understanding has involved: improved recognition of source areas and the potential for downstream effects of slope failure; improved understanding of hydrological conditions required for failure; and a general theory of slope stability in shallow colluvium, including the role of plants, fires, timber harvest, and other disturbances. The theory acknowledges spatial variability in topographic and geotechnical terrain characteristics, the stochastic nature of climatic triggering events such as forest fires and rainstorms, and the integrating nature of channel networks in modulating the cumulative effects of transient processes within a basin. Anthropogenic fire regimes, road effects, and timber harvest can readily be included. Continued application and modification of the theory over an expanded geographical range require improvements in field data and their systematic storage in spatial databases. Improvements in digital topographic data for mountain basins, systematic network-wide surveys of channel conditions, and new technology for rapid documentation of soil depths in landslide source areas would enhance the prediction of mass failure, its consequences for channel habitat, and the basin-wide or regional distribution of hillslope and channel conditions. Computations of the probabilities of transient effects throughout basins could then form the basis of ecological risk analyses. Large-scale spatial data sets of a few critical variables are required before this next level of understanding can be developed and applied to sedimentation impacts on ecosystems and other resources.     

Determining Relative Contributions of Vegetation and Topography to Burn Severity from LANDSAT Imagery

Fire is a dominant process in boreal forest landscapes and creates a spatial patch mosaic with different burn severities and age classes. Quantifying effects of vegetation and topography on burn severity provides a scientific basis on which forest fire management plans are developed to reduce catastrophic fires. However, the relative contribution of vegetation and topography to burn severity is highly debated especially under extreme weather conditions. In this study, we hypothesized that relationships of vegetation and topography to burn severity vary with fire size. We examined this hypothesis in a boreal forest landscape of northeastern China by computing the burn severity of 24 fire patches as the difference between the pre- and post-fire Normalized Difference Vegetation Index obtained from two Landsat TM images. The vegetation and topography to burn severity relationships were evaluated at three fire-size levels of small (<100 ha, n = 12), moderate (100–1,000 ha, n = 9), and large (>1,000 ha, n = 3). Our results showed that vegetation and topography to burn severity relationships were fire-size-dependent. The burn severity of small fires was primary controlled by vegetation conditions (e.g., understory cover), and the burn severity of large fires was strongly influenced by topographic conditions (e.g., elevation). For moderate fires, the relationships were complex and indistinguishable. Our results also indicated that the pattern trends of relative importance for both vegetation and topography factors were not dependent on fire size. Our study can help managers to design fire management plans according to vegetation characteristics that are found important in controlling burn severity and prioritize management locations based on the relative importance of vegetation and topography.     

Evaluation of the Fire Simulation Processes of the National Fire Management System's Initial Attack Analysis Processor

The fire simulation processes of the National Fire Management System's (NFMAS) Initial Attack Analysis (IAA) processor were evaluated by conducting two types of sensitivity analysis: one based on a hypothetical set of data to assess IAA's outputs under a wide range of fire input values, and the other using an actual Stanislaus National Forest database to test IAA's validity with a real set of data. The results revealed that IAA's outputs (projected annual number of fires and area burned) were most sensitive, in descending order, to the input values of the fire spread rate, the productivity rates of the suppression forces, and the initial attack time, for all fuel models tested. In contrast, IAA's outputs were extremely insensitive to variations in the fire size at discovery. Changes are necessary in the ways IAA incorporates the fire size at discovery to facilitate the comparison among various fire detection options. The program's “escaped fire situation” analysis was found inadequate, because the projected annual frequencies and final sizes of the simulated escaped fire events produced unacceptable results with the Stanislaus National Forest database. Assigning final sizes to simulated escaped fires according to the fire intensity level in which they are historically expected to occur provides a consistent way of calculation of the projected annual area burned and the consequent cost plus net value change (C + NVC).     

Human-caused wildfire risk rating for prevention planning in Spain

This paper identifies human factors associated with high forest fire risk in Spain and analyses the spatial distribution of fire occurrence in the country. The spatial units were 6,066 municipalities of the Spanish peninsular territory and Balearic Islands. The study covered a 13-year series of fire occurrence data. One hundred and eight variables were generated and input to a dedicated Geographic Information System (GIS) to model different factors related to fire ignition. After exploratory analysis, 29 were selected to build a predictive model of human fire ignition using logistic regression analysis. The binary model estimated the probability of high or low occurrence of forest fires, as defined by an ignition danger index that is currently used by the Spanish forest service (number of fires divided by forest area in each municipality). Thirteen explanatory variables were identified by the model. They were related to agricultural landscape fragmentation, agricultural abandonment and development processes. The prediction agreement found between the model binary outputs and the historical fire data was 85.3% for the model building dataset (60% of municipalities). A slightly lower predictive power (76.2%) was found for the validation data (the remaining 40%). The probabilistic output of the logistic was significantly related to the raw ignition index (Spearman correlation of 0.710) used by the Spanish Forest Service. Therefore, the model can be considered a good predictor of human-caused fire risk, aiding spatial decisions related to prevention planning in Spanish municipalities.     

Mapping probability of fire occurrence in San Jacinto Mountains,California, USA

An ecological data base for the San Jacinto Mountains, California, USA, was used to construct a probability model of wildland fire occurrence. The model incorporates both environmental and human factors, including vegetation, temperature, precipitation, human structures, and transportation. Spatial autocorrelation was examined for both fire activity and vegetation to determine the specification of neighborhood effects in the model. Parameters were estimated using stepwise logistic regressions. Among the explanatory variables, the variable that represents the neighborhood effects of spatial processes is shown to be of great importance in the distribution of wildland fires. An important implication of this result is that the management of wildland fires must take into consideration neighborhood effects in addition to environmental and human factors. The distribution of fire occurrence probability is more accurately mapped when the model incorporates the spatial term of neighborhood effects. The map of fire occurrence probability is useful for designing large-scale management strategies of wildfire prevention.     

SPATIAL CLUSTERS OF SUBSURFACE DRAINAGE WATER NO3‐N LEACHING LOSSES1

ABSTRACT: Grouping of nitrate‐nitrogen (NO₃‐N) leaching losses from agricultural fields into spatial clusters can help determine the cause/effect relationships for their occurrence. This study was designed to investigate the spatial relationships of low, medium, and high NO₃‐N leaching losses clusters with soil and landscape attributes using cluster and discriminant analysis and the map overlay capability of a geographical information system (GIS). Field measured data of a six‐year (1993 through 1998) study on NO₃‐N leaching losses from 36 experimental fields at the Iowa State University's northeastern research center near Nashua, Iowa, were normalized on an annual basis to compare over the years. The cluster analysis resulted in the formation of three clusters based on the satisfactory evaluation criteria of pseudo‐F statistic, cubic clustering criterion, and R² values. The discriminant analysis, carried out on the basis of clusters, identified elevation and subsurface drainage as the factors that contributed significantly (p > 0.01) in discriminating among these clusters. The verification of discriminant functions developed on these factors predicted the cluster membership for all the groups with an overall accuracy of 86 percent. The map overlay analyses of GIS showed that spatial occurrence of the clusters transporting high NO₃‐N leaching losses was affected by the interaction of soil type and elevation levels.     

A Statistical Sediment Yield Prediction Model Incorporating the Effect of Fires and Subsequent Storm Events1

Abstract: Alluvial fans are continuously being developed for residential, industrial, commercial, and agricultural uses in southern California. Development and alteration of alluvial fans need to consider the possibility of mud and debris flows from upstream mountain watersheds affected by fires. Accurate prediction of sediment yield (or hyper‐concentrated sediment yield) is essential for the design, operation, and maintenance of debris basins to safeguard properly the general populace. This paper presents a model for the prediction of sediment yields that result from a combination of fire and subsequent storm events. The watersheds used in this analysis are located in the foothills of the San Gabriel Mountains in southern California. A multiple regression analysis is first utilized to establish a fundamental statistical relationship for sediment yield as a function of relief ratio, drainage area, maximum 1‐h rainfall intensity and fire factor using 45 years of data (1938‐1983). In addition, a method for multi‐sequence sediment yield prediction under fire conditions was developed and calibrated using 17 years of sediment yield, fire, and precipitation data for the period 1984‐2000. After calibration, this model was verified by applying it to provide a prediction of the sediment yields for the 2001‐2002 fire events in southern California. The findings indicate a strong correlation between the estimated and measured sediment yields. The proposed method for sequence sediment yield prediction following fire events can be a useful tool to schedule cleanout operations for debris basins and to develop an emergency response strategy for the southern California region where plentiful sediment supplies exist and frequent fires occur.     

Socioeconomic backwardness and vulnerability to climate change: evidence from Uttar Pradesh state in India

This study tests the thesis that ‘vulnerability to climate change is not only a result of biophysical events alone but also influenced by the socioeconomic conditions in which climate change occurs’. The study chose Uttar Pradesh (UP), a state in India, for its importance in the nation's food and nutrition security programme and its high sensitivity to climate change. It uses an indicator approach to see which districts of UP are the most vulnerable to climate change, and attempts to identify the factors on a set of explanatory variables. The study finds that infrastructurally and economically developed districts are less vulnerable to climate change; in other words, vulnerability to climate change and variability is linked with social and economic development. This observation is corroborated by the findings of relational analysis wherein livestock, forestry, consumption of fertilizer, per capita income, and infant mortality rate are observed to be important correlates of vulnerability to climate change.     

Lightning Fires in a Brazilian Savanna National Park: Rethinking Management Strategies

Fire occurrences and their sources were monitored in Emas National Park, Brazil (17°49′–18°28′S; 52°39′–53°10′W) from June 1995 to May 1999. The extent of burned area and weather conditions were registered. Forty-five fires were recorded and mapped on a GIS during this study. Four fires occurred in the dry winter season (June–August; 7,942 ha burned), all caused by humans; 10 fires occurred in the seasonally transitional months (May and September) (33,386 ha burned); 31 fires occurred in the wet season, of which 30 were caused by lightning inside the park (29,326 ha burned), and one started outside the park (866 ha burned). Wet season lightning fires started in the open vegetation (wet field or grassy savanna) at a flat plateau, an area that showed significantly higher fire incidence. On average, winter fires burned larger areas and spread more quickly, compared to lightning fires, and fire suppression was necessary to extinguish them. Most lightning fires were patchy and extinguished primarily by rain. Lightning fires in the wet season, previously considered unimportant episodes, were shown to be very frequent and probably represent the natural fire pattern in the region. Lightning fires should be regarded as ecologically beneficial, as they create natural barriers to the spread of winter fires. The present fire management in the park is based on the burning of preventive firebreaks in the dry season and exclusion of any other fire. This policy does not take advantage of the beneficial effects of the natural fire regime and may in fact reduce biodiversity. The results presented here stress the need for reevaluating present policies and management procedures concerning fire in cerrado conservation areas.     

Mapping wildland-urban interfaces at large scales integrating housing density and vegetation aggregation for fire prevention in the South of France

Every year, more than 50,000 wildland fires affect about 500,000 ha of vegetation in southern European countries, particularly in wildland-urban interfaces (WUI). This paper presents a method to characterize and map WUIs at large scales and over large areas for wildland fire prevention in the South of France. Based on the combination of four types of building configuration and three classes of vegetation structure, 12 interface types were classified. Through spatial analysis, fire ignition density and burned area ratio were linked with the different types of WUI. Among WUI types, isolated WUIs with the lowest housing density represent the highest level of fire risk.