NBIC-ACS Stage 2 Forest Fire Danger Index - projected scenario, 20% and 10% annual exceedance probabilities

McArthur Forest Fire Danger Index (FFDI) is a fire weather potential index that describes how current weather conditions and recent precipitation patterns could support a fire in forested areas. FFDI calculations are based on McArthur, A.G. (1967) and dependent on air temperature, relative humidity, wind speed and drought factor. FFDI was originally implemented as a circular slide rule that constrained the returned index between 0 and 100, where 100 was representative of the worst Australian bushfire at that time, Victoria 1939. The data provided here is unbounded.\nHere we provide predicted upper-bound FFDI values across the Australian landscape, defined for a set of annual exceedance probabilities modelled using extreme values analysis on more than 43 years of hourly data.\nThe FFDI potential rasters represent reasonable worst case extreme conditions. Specifically, the rasters represent the FFDI potential for given AEPs of 20%, 10%, 5%, 2% and 1%. These FFDI AEPs are based on the projected weather timeseries developed by NBIC using the historical regional weather reanalysis dataset BARRA-R2 [1], and CMIP6-CCAM Regional Climate Models (RCM) [2].\nThe Regional Climate Models (RCM) considered are:\n•\tACCESS ESM 1.5\n•\tEC-Earth 3\n•\tCMCC ESM2\n•\tCNRM ESM 2.1\n•\tNCAR CESM2\n•\tNorESM2 MM\nThe future climate change scenarios considered are:\n•\tShared Socioeconomic Pathway (SSP) SSP 1-26: Sustainability \n•\tSSP 3-70: Regional Rivalry\n•\tSSP 3-70: Regional Rivalry, using wind speed from the baseline scenario \nThe combination of RCMs and future climate change scenarios results in a suite of 18 projected FFDI potential datasets.\nLineage: The McArthur Forest Fire Danger Index (FFDI) data - projected scenario is calculated using the CMIP6-ERA Regional Climate Model data produced with CCAM by CSIRO (https://dx.doi.org/10.25914/rd73-4m38). Variables used are near surface air temperature, precipitation, near surface relative humidity and near surface wind speed.\nThose inputs are combined using the equations developed by Noble et al. (1980) [1] based on the work by McArthur (1967) [2]. McArthur Drought Factor is calculated using the modified formulation published in Finkele et al. (2006) [3].\nFFDI is calculated for more than 43 years of hourly weather data and then processed using extreme values analysis to model the expected FFDI values at annual exceedance probabilities.\n[1] Noble, I.R., Gill, A.M. and Bary, G.A.V., 1980. McArthur's fire‐danger meters expressed as equations. Australian journal of ecology, 5(2), pp.201-203. https://doi.org/10.1111/j.1442-9993.1980.tb01243.x\n[2] McArthur, A.G., 1967. Fire behaviour in eucalypt forests: Forestry and Timber Bureau Leaflet 107. Forestry and Timber Bureau: Canberra, Australia.\n[3] Finkele, K., Mills, G.A., Beard, G. and Jones, D.A., 2006. National gridded drought factors and comparison of two soil moisture deficit formulations used in prediction of Forest Fire Danger Index in Australia. Australian Meteorological Magazine, 55(3), pp.183-197.

麦克阿瑟森林火险指数（McArthur Forest Fire Danger Index, FFDI）是一类火险气象潜力指数，用于描述当前气象条件与近期降水格局对林区火灾的潜在支撑作用。FFDI的计算依据McArthur AG在1967年的研究，依赖于气温、相对湿度、风速与干旱因子。该指数最初以圆形计算尺的形式实现，将输出指数限定在0至100之间，其中100代表当时澳大利亚最严重的丛林火灾——1939年维多利亚州大火。本次提供的数据集无此取值限制。

本次研究提供了澳大利亚全域的预测上限FFDI值，该数据集基于超过43年的逐时气象数据，通过极值分析建模得到了一系列年超越概率对应的FFDI值。

FFDI潜力栅格数据代表合理最坏情形极端条件。具体而言，这些栅格数据对应20%、10%、5%、2%与1%的给定年超越概率下的FFDI潜力值。此类基于年超越概率的FFDI数据，由NBIC基于历史区域气象再分析数据集BARRA-R2[1]与CMIP6-CCAM区域气候模式（Regional Climate Models, RCM）[2]生成的预测气象时间序列构建而来。

本次研究所采用的区域气候模式（RCM）包括：
• ACCESS ESM 1.5
• EC-Earth 3
• CMCC ESM2
• CNRM ESM 2.1
• NCAR CESM2
• NorESM2 MM

本次研究所考虑的未来气候变化情景包括：
• 共享社会经济路径（Shared Socioeconomic Pathway, SSP）SSP1-2.6：可持续发展情景
• SSP3-7.0：区域竞争情景
• SSP3-7.0：区域竞争情景（采用基准情景的风速数据）

区域气候模式与未来气候变化情景的组合共生成18套预测FFDI潜力数据集。

数据溯源：本次预测情景下的麦克阿瑟森林火险指数（FFDI）数据，基于CSIRO利用CCAM生成的CMIP6-ERA区域气候模式数据计算得到（https://dx.doi.org/10.25914/rd73-4m38）。所用变量包括近地面气温、降水、近地面相对湿度与近地面风速。
上述输入变量通过Noble等人1980年提出的公式进行组合，该公式基于McArthur 1967年的研究成果[2]。麦克阿瑟干旱因子则采用Finkele等人2006年发表的修正公式进行计算[3]。
研究首先基于超过43年的逐时气象数据计算FFDI，随后通过极值分析对年超越概率下的FFDI预期值进行建模。

[1] Noble IR, Gill AM, Bary GAV. 1980. 以公式形式表达的麦克阿瑟火险计. 澳大利亚生态学杂志, 5(2): 201-203. https://doi.org/10.1111/j.1442-9993.1980.tb01243.x
[2] McArthur AG. 1967. 桉树林火灾行为：林业与木材局手册107. 澳大利亚堪培拉：林业与木材局.
[3] Finkele K, Mills GA, Beard G, Jones DA. 2006. 澳大利亚全国网格化干旱因子及两种用于森林火险指数预测的土壤水分亏缺公式对比. 澳大利亚气象杂志, 55(3): 183-197.