Habitat condition time-series for the Cooper region

\n## **Abstract** \nThis dataset represents a historic annual time-series of estimated habitat condition across the study region, from 2001 to 2018 inclusive. These spatial grids were developed by combining two spatial data products: (1) a new analytical approach called Compere, which contrasts the vegetation cover of each location with a set of environmentally similar locations across the region, for each year; (2) a temporally static spatial layer on the total length of all linear anthropogenic disturbances, in each 500 m grid cell across the study region. The resultant habitat condition spatial time-series is therefore intended to combine information on both localised (e.g. roads, seismic surveys) and dispersed (e.g grazing, fire) influences on the habitat condition for biodiversity of each location, ranging continuously from '0' (completely degraded) to '1' (pristine).\n\n\n## **Attribution** \nGeological and Bioregional Assessment Program\n\n\n## **History** \nTo derive a past-to-present time-series of habitat condition across the buffered study regions, we combined two spatial data products. The first set of spatial data come from a new analytical approach called Compere, which contrasts the vegetation cover of each location with a set of environmentally similar locations across the region. For the present purpose, we applied the Absolute Range Ratio (ARR) metric from Compere, which for each location at a time-point is the observed vegetation cover divided by the maximum vegetation cover across all the environmentally similar locations at that time-point. The ARR spatial layers were available for each year from 2001 to 2018 inclusive. To translate the ARR spatial layers to better represent a habitat condition metric (h_ARR), we rescaled the ARR values as: \r\nh_ARR=ARR+(k×((100-ARR)/100))\r\nwith the scalar k specifying the minimum habitat condition value set to 40 % for the present analysis.\r\n\r\nThe second spatial dataset used in deriving habitat condition for biodiversity was a spatial layer on the total length of all linear disturbances in each 500 m grid cell across the study regions. We converted this layer into a habitat condition metric (hL) by assuming complete habitat loss for a width of 10 m for all linear disturbances, then taking the inverse of the proportion of each grid cell area that was disturbed.\r\n\r\nThese two data sources on habitat condition provide complementary information. The h_ARR condition metric derived from the Compere analysis is useful in detecting the broadscale impacts of actions such as grazing and fire management across the region on habitat condition. In contrast, the hL condition metric derived from the spatial data on linear disturbances is useful in identifying known disturbances, such as roads, fence-lines and seismic survey lines. We therefore combined these two condition metrics, using a conservative approach of taking the minimum condition value from each of these metrics, for each grid cell:\r\nh=min(h_ARR,h_L )\r\nassuming a constant level of linear disturbance over the time period of the Compere analysis (2001-2018). This provided an historic annual time series of habitat condition across each study region.

## **摘要**
本数据集涵盖研究区域2001年至2018年（含首尾年份）的历史性年度生境状况估算时间序列。该空间栅格数据集由两类空间数据产品融合构建：(1) 一种名为Compere的新型分析方法，该方法针对每一年度，将各点位的植被覆盖情况与区域内一组环境相似的点位进行对比；(2) 研究区域内每个500米栅格单元中所有线性人为干扰总长度的时间静态空间图层。最终生成的生境状况空间时间序列，旨在整合局地性（如道路、地震勘探）与分散性（如放牧、火灾）两类因素对各点位生物多样性生境状况的影响，其取值范围连续覆盖0（完全退化）至1（原始完好）。

## **数据归属**
地质与生物区域评估项目

## **数据构建历程**
为生成缓冲研究区域内从过去至当前的生境状况时间序列，我们融合了两类空间数据产品。第一类空间数据源自名为Compere的新型分析方法，该方法将每个点位的植被覆盖与区域内一组环境相似的点位进行对比。针对本研究目标，我们采用了Compere方法中的绝对范围比（Absolute Range Ratio, ARR）指标：对于某一时间点的每个点位，ARR为该点位观测到的植被覆盖除以同期所有环境相似点位的最大植被覆盖。2001年至2018年各年度的ARR空间图层均已公开可用。为将ARR空间图层转换为更贴合生境状况的指标（记为h_ARR），我们对ARR值进行了如下重缩放：
h_ARR=ARR+(k×((100-ARR)/100))
其中标量k为本分析中设定的最小生境状况值，取值为40%。

第二类用于推导生物多样性生境状况的空间数据集，为研究区域内每个500米栅格单元中所有线性干扰总长度的空间图层。我们将该图层转换为生境状况指标（记为hL）：假设所有线性干扰的影响宽度为10米，即造成完全生境丧失，随后计算每个栅格单元中未受干扰区域占比的倒数，以此作为hL。

上述两类生境状况数据源具备信息互补性。由Compere分析得到的h_ARR状况指标，可有效检测区域内放牧、火灾管理等行动对生境状况产生的大范围影响；相较而言，由线性干扰空间数据推导得到的hL状况指标，可精准识别道路、围栏线、地震勘探线等已知干扰源。因此，我们采用保守的融合策略，对每个栅格单元取两类指标中的最小值，以整合这两类生境状况指标：
h=min(h_ARR,h_L)
本分析假设2001-2018年期间线性干扰水平保持恒定，最终得到各研究区域的历史性年度生境状况时间序列。