Geographically Weighted Negative Binomial Regression Model Predicts Wildfire Occurrence in the Great Xing’an Mountains Better Than Negative Binomial Model

Wildfire is a major disturbance that affects large area globally every year. Thus, a better prediction of the likelihood of wildfire occurrence is essential to develop appropriate fire prevention measures. We applied a global negative Binomial (NB) and a geographically weighted negative Binomial regression (GWNBR) models to determine the relationship between wildfire occurrence and its drivers factors in the boreal forests of the Great Xing’an Mountains, northeast China. Using geo-weighted techniques to consider the geospatial information of meteorological, topographic, vegetation type and human factors, we aimed to verify whether the performance of the NB model can be improved. Our results confirmed that the model fitting and predictions of GWNBR model were better than the global NB model, produced more precise and stable model parameter estimation, yielded a more realistic spatial distribution of model predictions, and provided the detection of the impact hotpots of these predictor variables. We found slope, vegetation cover, average precipitation, average temperature, and average relative humidity as important predictors of wildfire occurrence in the Great Xing’an Mountains. Thus, spatially differing relations improves the explanatory power of the global NB model, which does not explain sufficiently the relationship between wildfire occurrence and its drivers. Thus, the GWNBR model can complement the global NB model in overcoming the issue of nonstationary variables, thereby enabling a better prediction of the occurrence of wildfires in large geographical areas and improving management practices of wildfire.

Distribution and Driving Factors of Forest Swamp Conversions in a Cold Temperate Region

Forest swamps are widely distributed in cold temperate regions, with important landscape and ecological functions. They are prone to conversion caused by complex factors. Forest swamp conversions involve forest swamping, meadow swamping, water body swamping, and conversion to farmland. An understanding of the landscape characteristics and primary environmental factors driving forest swamp conversions is imperative for exploring the mechanism of forest swamp conversions. We investigated the landscape characteristics of forest swamp conversions and quantified the relative importance of environmental factors driving these conversions for the period from 1990 to 2015 in the Great Xing’an Mountains of China. We found that forest swamping displayed high patch numbers (34,916) and density (8.51/100 ha), commonly occurring at the edge of large areas of forests. Meadow swamping was localized with low patch numbers (3613) and density (0.88/100 ha) due to lack of water recharge from ground water. Water body swamping had complex shapes (perimeter area ratio mean = 348.32) because of water table fluctuations and helophyte growth during this conversion process. Conversions to farmland presented fairly regular (perimeter area ratio mean = 289.91) and aggregated (aggregation index = 67.82) characteristics affected by agricultural irrigation and management. We found that climatic and geomorphic factors were relatively important compared to topographic factors for forest swamp conversions. Negative geomorphic conditions provided the waterlogging environment as a precondition of swamp formation. Sufficient precipitation was an important source of water recharge due to the existence of permafrost regions and long-term low temperature reduced the evaporation of swamps water and the decomposition rate of organisms. These wet and cold climatic conditions promoted forest swamp development in cold temperate regions. Humans exerted a relatively important role in forest swamping and conversions to farmland. Fire disturbance and logging accelerated the conversion from forest to swamp. This study provides scientific information necessary for the management and conservation of forest swamp resources in cold temperate regions.

Evaluating the applicability of predicting dead fine fuel moisture based on the hourly Fine Fuel Moisture Code in the south-eastern Xing’an Mountains of China

To evaluate the applicability of the hourly Fine Fuel Moisture Code (FFMC) to the south-eastern Great Xing’an Mountains, dead fine fuel moisture (Mf) was observed under less-sheltered and sheltered conditions in Scots pine (Pinus sylvestris var. mongolica), larch (Larix gmelinii) and oak (Quercus mongolicus) stands during the summer and autumn of 2014. Standard FFMC and locally calibrated FFMC values calculated hourly were tested using Mf observations and weather data, and the results showed that the Mf loss rate in the less-sheltered forest floor was markedly higher than that in the sheltered forest floor (P < 0.05). The standard hourly FFMC underestimated Mf, especially in stands of larch, the dominant species in the Great Xing’an Mountains, and Mf for rainy days in Scots pine and oak stands. However, the calibrated hourly FFMC predicted Mf in all three forest stands very well (R2 ranged from 0.920 to 0.969; mean absolute errorfrom 2.93 to 6.93, and root-mean-squared errorfrom 4.09 to 7.87), which suggested that it was sufficiently robust for those stands around the observation period. This study will improve the accuracy of Mf predictions to aid fire control efforts in the Great Xing’an Mountains and provide a basis for hourly FFMC model calibration.