Girardin Martin P, Mudelsee Manfred
Natural Resources Canada, Canadian Forest Service, Laurentian Forestry Centre, 1055 du P.E.P.S., P.O. Box 10380, Stn. Sainte-Foy, Quebec G1V4C7, Canada.
Ecol Appl. 2008 Mar;18(2):391-406. doi: 10.1890/07-0747.1.
Climate change in Canadian boreal forests is usually associated with increased drought severity and fire activity. However, future fire activity could well be within the range of values experienced during the preindustrial period. In this study, we contrast 21st century forecasts of fire occurrence (FireOcc, number of large forest fires per year) in the southern part of the Boreal Shield, Canada, with the historical range of the past 240 years statistically reconstructed from tree-ring width data. First, a historical relationship between drought indices and FireOcc is developed over the calibration period 1959-1998. Next, together with seven tree-ring based drought reconstructions covering the last 240 years and simulations from the CGCM3 and ECHAM4 global climate models, the calibration model is used to estimate past (prior to 1959) and future (post 1999) FireOcc. Last, time-dependent changes in mean FireOcc and in the occurrence rate of extreme fire years are evaluated with the aid of advanced methods of statistical time series analysis. Results suggest that the increase in precipitation projected toward the end of the 21st century will be insufficient to compensate for increasing temperatures and will be insufficient to maintain potential evapotranspiration at current levels. Limited moisture availability would cause FireOcc to increase as well. But will future FireOcc exceed its historical range? The results obtained from our approach suggest high probabilities of seeing future FireOcc reach the upper limit of the historical range. Predictions, which are essentially weighed on northwestern Ontario and eastern boreal Manitoba, indicate that, by 2061-2100, typical FireOcc could increase by more than 34% when compared with the past two centuries. Increases in fire activity as projected by this study could negatively affect the implementation in the next century of forest management inspired by historical or natural disturbance dynamics. This approach is indeed feasible only if current and future fire activities are sufficiently low compared with the preindustrial fire activity, so a substitution of fire by forest management could occur without elevating the overall frequency of disturbance. Conceivable management options will likely have to be directed toward minimizing the adverse impacts of the increasing fire activity.
加拿大北方森林的气候变化通常与干旱程度加剧和火灾活动增加有关。然而，未来的火灾活动很可能处于工业化前时期经历的数值范围内。在本研究中，我们将加拿大北方盾形区南部21世纪的火灾发生预测（FireOcc，每年大型森林火灾的数量）与根据树轮宽度数据统计重建的过去240年的历史范围进行对比。首先，在1959 - 1998年的校准期内建立干旱指数与FireOcc之间的历史关系。接下来，结合涵盖过去240年的七个基于树轮的干旱重建数据以及CGCM3和ECHAM4全球气候模型的模拟结果，使用校准模型来估计过去（1959年之前）和未来（1999年之后）的FireOcc。最后，借助统计时间序列分析的先进方法，评估平均FireOcc和极端火灾年份发生率随时间的变化。结果表明，预计到21世纪末降水量的增加将不足以补偿气温的上升，也不足以将潜在蒸散量维持在当前水平。有限的水分供应也会导致FireOcc增加。但是未来的FireOcc会超过其历史范围吗？我们的方法得出的结果表明，未来FireOcc很有可能达到历史范围的上限。主要针对安大略省西北部和马尼托巴省东部北方地区的预测表明，到2061 - 2100年，与过去两个世纪相比，典型的FireOcc可能会增加超过34%。本研究预测的火灾活动增加可能会对下个世纪受历史或自然干扰动态启发的森林管理实施产生负面影响。只有当前和未来的火灾活动与工业化前的火灾活动相比足够低，这种方法才切实可行，这样森林管理才能替代火灾而不会提高总体干扰频率。可以想象，管理选项可能必须旨在将火灾活动增加的不利影响降至最低。
