Key message: We have explored the impacts of forest thinning on wildland fire behavior using a process based model. Simulating different degrees of thinning, we found out that forest thinning should be conducted cautiously as there could be a wide range of outcomes depending upon the post-thinning states of fuel availability, fuel connectivity, fuel moisture and micrometeorological features such as wind speed. Context: There are conflicting reports in the literature regarding the effectiveness of forest thinning. Some studies have found that thinning reduces fire severity, while some studies have found that thinning might lead to enhanced fire severity. Aims: Our goal was to evaluate if both of these outcomes are possible post thinning operations and what are the limiting conditions for post thinning fire behavior. Methods: We used a process based model to simulate different degrees of thinning systematically, under two different conditions, where the canopy fuel moisture was unchanged and when the canopy fuel moisture was also depleted post thinning. Both of these scenarios are reported in the literature. Results: We found out that a low degree of thinning can indeed increase fire intensity, especially if the canopy fuel moisture is low. A high degree of thinning was effective in reducing fire intensity. However, thinning also increased rate of spread under some conditions. Interestingly, both intensity and rate of spread were dependent on the competing effects of increased wind speed, fuel loading and canopy fuel moisture. Conclusion: We were able to find the limits of fire behavior post thinning and actual fire behavior is likely to be somewhere in the middle of the theoretical extremes explored in this work. The actual fire behavior post thinning should depend on the site specific conditions which would determine the outcome of the interplay among the aforementioned conditions. The work also highlights that policymakers should be careful about fine scale canopy architectural attributes and micrometeorological aspects when planning fuel treatment operations.

Geographic Information Systems (GIS) and related technologies have generated substantial interest among statisticians with regard to scalable methodologies for analyzing large spatial datasets. A variety of scalable spatial process models have been proposed that can be easily embedded within a hierarchical modeling framework to carry out Bayesian inference. While the focus of statistical research has mostly been directed toward innovative and more complex model development, relatively limited attention has been accorded to approaches for easily implementable scalable hierarchical models for the practicing scientist or spatial analyst. This article discusses how point-referenced spatial process models can be cast as a conjugate Bayesian linear regression that can rapidly deliver inference on spatial processes. The approach allows exact sampling directly (avoids iterative algorithms such as Markov chain Monte Carlo) from the joint posterior distribution of regression parameters, the latent process and the predictive random variables, and can be easily implemented on statistical programming environments such as R.

With the growing capabilities of Geographic Information Systems(GIS) and user-friendly software, statisticians today routinely encounter geographicallyreferenced data containing observations from a large number of spatial locationsand time points. Over the last decade, hierarchical spatiotemporal processmodels have become widely deployed statistical tools for researchers to better understandthe complex nature of spatial and temporal variability. However, fittinghierarchical spatiotemporal models often involves expensive matrix computationswith complexity increasing in cubic order for the number of spatial locations andtemporal points. This renders such models unfeasible for large data sets. Thisarticle offers a focused review of two methods for constructing well-defined highlyscalable spatiotemporal stochastic processes. Both these processes can be used as“priors” for spatiotemporal random fields. The first approach constructs a lowrankprocess operating on a lower-dimensional subspace. The second approachconstructs a Nearest-Neighbor Gaussian Process (NNGP) that ensures sparseprecision matrices for its finite realizations. Both processes can be exploited asa scalable prior embedded within a rich hierarchical modeling framework to deliverfull Bayesian inference. These approaches can be described as model-basedsolutions for big spatiotemporal datasets. The models ensure that the algorithmiccomplexity has ∼ n floating point operations (flops), where n the number of spatiallocations (per iteration). We compare these methods and provide some insightinto their methodological underpinnings.