Fire-dependent plant species persist in the humid North America eastern forests, which is devoid of a wet or dry season. But fire suppression has altered vegetation distributions. Knowledge of the natural fire regime can provide forest managers with the information needed for the conservation of these species. This thesis seeks to understand lightning-ignited fires in the Grandfather Ranger District of Pisgah National Forest in western North Carolina. To this end, I used self-organized maps, an automated synoptic typing procedure, to discern atmospheric circulation patterns. Those atmospheric circulation patterns were then associated with lightning-ignited fires and with lightning strikes to understand the atmospheric configurations that favor these events. I then detailed the surface climatic variables and synoptic classifications associated with each individual lightning-ignition. Finally I used point-pattern analysis to discern the association between lightning-ignitions and mountain golden heather (Hudsonia montana), a fire-dependent endemic shrub. A twenty year period was selected for this study. In Objective 1, 500mb geopotential heights, using North American Regional Reanalysis (NARR) data, were used to create synoptic types via self-organized maps. A 12 type composite was selected for this analysis. Those patterns were then associated with lightning ignited fires using National Interagency Fire Management Integrated Database (NIFMID) data. There were 39 lightning-ignited fire events. Lightning-ignitions occurred more than expected in synoptic types associated with high pressure, and predominantly in June. Objective 2 created a lightning climatology. Lightning strikes occurred most often during summer, peaking in July. Lightning strikes associated more than expected with synoptic types with high pressure zonal flow with transitional gradients over the study area. Objective 3 detailed the lightning-ignition events. The majority of events experienced high pressure before, with dry cold frontal movement the day before ignition, and high pressure after ignition. Objective 4 used Ripley's K analysis and bivariate analysis to discern the spatial arrangement of lightning-ignitions and Hudsonia montana. I found lightning-ignitions to be random at small scales and aggregated at larger scales, and Hudsonia montana to be aggregated across the landscape. When compared using bivariate analysis I found lightning-ignition locations and Hudsonia montana populations to be random at small scales and aggregated at larger scales. This suggest lightning-ignitions and Hudsonia montana populations are not associated with each other by point location, and lightning-ignited fires would need to burn large areas to impact the plant species.
Fire-dependent plant species persist in the humid North America eastern forests, which is devoid of a wet or dry season. But fire suppression has altered vegetation distributions. Knowledge of the natural fire regime can provide forest managers with the information needed for the conservation of these species. This thesis seeks to understand lightning-ignited fires in the Grandfather Ranger District of Pisgah National Forest in western North Carolina. To this end, I used self-organized maps, an automated synoptic typing procedure, to discern atmospheric circulation patterns. Those atmospheric circulation patterns were then associated with lightning-ignited fires and with lightning strikes to understand the atmospheric configurations that favor these events. I then detailed the surface climatic variables and synoptic classifications associated with each individual lightning-ignition. Finally I used point-pattern analysis to discern the association between lightning-ignitions and mountain golden heather (Hudsonia montana), a fire-dependent endemic shrub.
A twenty year period was selected for this study. In Objective 1, 500mb geopotential heights, using North American Regional Reanalysis (NARR) data, were used to create synoptic types via self-organized maps. A 12 type composite was selected for this analysis. Those patterns were then associated with lightning ignited fires using National Interagency Fire Management Integrated Database (NIFMID) data. There were 39 lightning-ignited fire events. Lightning-ignitions occurred more than expected in synoptic types associated with high pressure, and predominantly in June. Objective 2 created a lightning climatology. Lightning strikes occurred most often during summer, peaking in July. Lightning strikes associated more than expected with synoptic types with high pressure zonal flow with transitional gradients over the study area. Objective 3 detailed the lightning-ignition events. The majority of events experienced high pressure before, with dry cold frontal movement the day before ignition, and high pressure after ignition. Objective 4 used Ripley's K analysis and bivariate analysis to discern the spatial arrangement of lightning-ignitions and Hudsonia montana. I found lightning-ignitions to be random at small scales and aggregated at larger scales, and Hudsonia montana to be aggregated across the landscape. When compared using bivariate analysis I found lightning-ignition locations and Hudsonia montana populations to be random at small scales and aggregated at larger scales. This suggest lightning-ignitions and Hudsonia montana populations are not associated with each other by point location, and lightning-ignited fires would need to burn large areas to impact the plant species.
