Prescribed burning is an important management tool for the restoration and maintenance of tallgrass prairies. To improve fire behaviour prediction in tallgrass prairies, I assessed three different aspects of fire behaviour - heat of combustion, fuel load and rate of spread. Heat of combustion was found to vary amongst certain tallgrass species but the relatively small differences in means is unlikely to contribute significantly to fire behaviour. Average fuel loads in Ontario tallgrass prairie sites were found to be higher than current default value used in fire behaviour prediction. Three rapid fuel load assessment techniques were tested. Finally, the predictions of three fire behaviour prediction systems - the FBP System, BehavePlus and an Australian grassfire spread model, were compared with actual fire behaviour observations. The FBP System was found to perform poorly while both BehavePlus and the Australian model exhibited relatively strong relationships between observed and predicted rates of spread.

Prescribed burning is an important management tool for the restoration and maintenance of tallgrass prairies. To improve fire behaviour prediction in tallgrass prairies, I assessed three different aspects of fire behaviour - heat of combustion, fuel load and rate of spread. Heat of combustion was found to vary amongst certain tallgrass species but the relatively small differences in means is unlikely to contribute significantly to fire behaviour. Average fuel loads in Ontario tallgrass prairie sites were found to be higher than current default value used in fire behaviour prediction. Three rapid fuel load assessment techniques were tested. Finally, the predictions of three fire behaviour prediction systems - the FBP System, BehavePlus and an Australian grassfire spread model, were compared with actual fire behaviour observations. The FBP System was found to perform poorly while both BehavePlus and the Australian model exhibited relatively strong relationships between observed and predicted rates of spread.

Gräsbränder är ett återkommande problem i hela Sverige och kan påverka både egendom och människor. Det finns flera faktorer som påverkar uppkomsten av gräsbränder. Det kan vara väder/klimat, mänsklig påverkan, typ av gräs och antändningsmedel. I denna studie är målet att kartlägga vart det sker flest gräsbränder i Sverige, varför de sker och när säsongen startar och slutar. Ett annat mål med studien är att undersöka hur utsatta kommuner ser på och hanterar problemet med gräsbränder. Slutligen var syftet också att undersöka om framtida klimatförändringar kan komma att påverka gräsbrandsituationen i Sverige. För att svara på studiens mål valdes att använda en statistisk undersökning som bygger på insatsstatistik från Myndigheten för samhällsskydd och beredskap (MSB). Det genomfördes även två intervjuer med kommunanställda, en från en kommun i södra Sverige och en från en kommun i norra Sverige. Intervjudata analyserades utifrån en innehållsanalys. Slutligen användes även vetenskaplig litteratur  och studier av klimatscenarier för att undersöka klimatförändringars roll för vegetationsbränder.  Resultaten av studien visar att det är Göteborg som är den kommun där flest gräsbränder sker, därefter kommer Stockholm, Halmstad och Borås. Västra Götaland är det län som drabbas mest frekvent av gräsbränder i Sverige och även det län som störst gräsbränder sker. Det är däremot Halland som har flest gräsbränder sett till antal bränder per capita Den vanligaste kända orsaken till gräsbränder är eldning av gräs men det finns också en stor andel där orsaken är okänd. Vid intervjuerna framkom att gräsbränder anses som ett litet problem och att kommunernas sätt att hantera dem är bra. Däremot kan ett problem med gräsbränder anses vara att det kan ta tid från andra räddningsuppdrag som kommunen har. Intervjuerna visar även att det finns en potential för att via utbildning och information till invånare minska antalet gräsbränder i kommunerna. Klimatförändringar kan komma att påverka gräsbrandsutvecklingen i Sverige. Om luftfuktigheten kommer att öka, vilket scenarier visar, anses gräsbrandsrisken minska i form av färre branddagar, men det kan dock bli intensivare gräsbränder på grund av en längre växtsäsong och därmed en större mängde bränsle. / Grass fires are a recurrent problem throughout Sweden and can affect both property and people. There are factors that influence the emergence of grass fires. It could be the weather / climate, human impact, the type of grass and ignition agents. In this study, the goal is to identify where most grass fires occur in Sweden, why they occur, and when the season starts and ends. Another goal of the study is to examine how vulnerable municipalities manage and look at the problem of grass fires. Finally, the goal was also to investigate whether future climate change may affect the grass fire situation in Sweden. To answer the purpose of the study a statistical survey was made based on emergency response data from the Swedish Civil Contingencies Agency (MSB). There were also two interviews made with municipal employees, one from a municipality in southern Sweden and one from a municipality in northern Sweden. The interview data were analyzed with content analysis. Finally, research literature and scenario studies were used to study the role of climate change for vegetation fires.  The results of the study show that Gothenburg is the municipality where most grass fires occur, followed by Stockholm, Halmstad and Borås. Västra Götaland is the county having most frequent grass fires in Sweden, and is also the county where the greatest grass fires occur. Is, however, the county of Halland which has most grass fires, seen as fires per capita. The most common known cause of grass fires is burning of grass, but there is also a large proportion of fires for which the cause is unknown.  The interviews revealed that grass fires are considered a minor problem and that the municipalities’ way of dealing with them is good. However, one problem with grass fires, considered by the municipalities is that they may take time and resources away from other rescue missions. The interviews also showed that there is a potential for teaching of residents, as a way to reduce the number of grass fires in the municipalities. Climate change may affect grass fire development in Sweden. If humidity will increase grass fire risk is considered to decrease because of less frequent fire-days, but grass fires may become more intense due to longer growth season and therefore a great amount of fuel.

Climate change and exotic plant invasions are significant anthropogenic threats to desert community structure and resilience . In the Mojave Desert, the invasive grass red brome (Bromusrubens L) is increasing fire frequency and extent in response to climatic factors. The resilience of this ecosystem will be affected by how plant and animal communities respond to fire. To better understand these dynamics, we studied the environmental factors underlying changes in invasive grass fire regimes in the Mojave Desert and its structural and functional effects on plant and animal communities. Following fire, reestablishment of native vegetation can be preempted by repeated burning associated with the abundant exotic grass red brome. Red brome density is correlated with various climate and landscape variables, but to establish causality, we experimentally assessed germination and growth of red brome. Red brome responded positively to fall precipitation, finer-textured soils, fertile-islands soils, and soils from burned landscapes. Red brome germination is maximized in wet fall periods when adequate water and optimal temperatures overlap . To evaluate landscape responses of pre- and post-fire plant communities and the potential for repeated burning we analyzed vegetation greenness (NDVI) data from 1985-2011 in response to temperature and precipitation. Landscape analysis indicated that the dominance of exotic grasses increases on post-fire landscapes. Following wet fall and winter seasons, high red brome productivity increases fire potential. Without mitigation, the establishment of an invasive-plant-driven fire regime is likely and may drive state transitions from arid shrublands to arid annual grasslands. Potential revegetation of post-fire landscapes will depend at least in part upon the physiological response of surviving vegetation to post-fire landscapes. Plant physiological responses to post-fire landscapes were generally neutral or positive, suggesting that revegetation of post-fire landscapes is not precluded by resource loss associated with fire and may even be enhanced by post-fire conditions. This will likely translate to increased reproductive potential of surviving plants. Alterations to small mammal populations will likely play a role in the reestablishment of vegetation (both native and exotics) as small mammals have strong top-down effects in arid ecosystems. Diversity and species richness responded negatively to burned landscapes as Merriam's kangaroo rat (Dipodomys merriami) increased in abundance while other species practically disappeared from burned landscapes. Merriam's kangaroo rat affects propagule sources through direct consumption, and seed dispersal. Increases in abundance and dominance of Merriam's kangaroo rat will likely alter plant recruitment.

Results are presented from the 2021 test run of active fire detection using the Visual Infrared Imaging Radiometer Suite (VIIRS) instrument, that is currently onboard the polar satellites Suomi-NPP and NOAA-20. The test is performed by the Swedish Civil Contingencies Agency and the Swedish Meteorological and Hydrological Institute, in cooperation with local fire departments in Sweden. The aim of this report was to study the impact of clouds on the ability of active fire detection, as well as to identify objects that potentially can cause commission errors in the VIIRS 375 m active fire algorithm (false positive notifications). Also, the study aimed to investigate what may cause omission errors in the algorithm, and to show to what extent the detections can be used to represent the true time development of the wildfire front and burned area. Using a cloudmask and a cloudtype classification product from Nowcasting Satellite Application Facilities (NWC SAF), the impact of clouds was anlalyzed by comparing the cloud data with the obtained fire notifications from the satellites. Active fires and newly burned areas were also studied using Sentinel-2 imagery, specifically the False Color Urban and the Short Wave Infrared (SWIR) RGB composites, as well as images from the 842 nm band, making use of the relatively high spatial resolution as well as the spectral signatures of fire and newly burned vegetation. Detection of active fires occurred in both cloud free and completely cloud covered conditions. How-ever, roughly 70% of the detected vegetation fire pixels were obtained in conditions with 20% clouds or less. / I rapporten presenteras resultat från 2021 års test av satellitdetektering av skogs- och vegetationsbränder, ett test som genomförs av MSB och SMHI i samverkan med kommunala räddningstjänster. De två satelliter som ingår i testet (Suomi-NPP och NOAA-20) går i polära omloppsbanor och är utrustade med instrumentet Visible Infrared Imaging Radiometer Suite (VIIRS). Detta projekt syftade till att undersöka hur molnighet påverkar möjligheten till detektion med satellit, vilka objekt som potentiellt kan ge upphov till falska detektioner samt vad som kan orsaka uteblivna satellitdetektioner. Ytterligare ett mål med rapporten var att genom fallstudier av större bränder undersöka i vilken utsträckning satellitdetektionerna kan användas för att representera brandfrontens utveckling med tiden och brandens faktiska utbredning. Vid studierna av molnighet analyserades en molnmask och en molnklassificeringsprodukt från Nowcasting Satellite Application Facilities (NWC SAF). I utvärderingen användes även data från Sentinel-2 för att studera pågående bränder och avbränd yta, som syns tydligt i RGB-kompositerna False Color Urban och Short Wave Infrared (SWIR) och i 842 nm-bandet, tack vare den relativt höga bildupplösningen och den nyligen avbrända ytans spektralsignatur. Brand detekterades i både molnfria och helt molntäckta förhållanden. Drygt 70 % av detektionerna vid vegetationsbrand kom emellertid i förhållanden med 20 % moln eller mindre.