Pre-European estimate of mean annual concentration of dissolved phosphorus in soil water (PMnlConc0.Base)

The dissolved phosphorous (P) concentration is calculated as the ratio of the plant-available mineral P store to soil water store. Rainfall has a strong influence on the pattern of both these stores, but not on their ratio. The pattern of dissolved P concentrations is explained by the effect of air dryness (saturation deficit) on the dissolved P store (mainly controlled by NPP) which is greater than the effect of air dryness on the soil water store (mainly determined by rainfall or energy limitation). This gives a pattern of P concentration in soil water that decreases as the climate-average air dryness increases from temperate to semi-arid tropical environments. Derived from the BiosEquil model by Raupach et al. (2001a; 2001b).

Soil nutrient outputs of the BiosEquil model
Nutrient status is one of the key limiting factors determining the productivity of Australian vegetation systems, but is only broadly represented by gross nutrient status an attribute compiled from the Atlas of Australian Soils (McKenzie et al. 2000). We therefore additionally compiled the 0.05°gridded pre-European (base) predictions of carbon, nitrogen and phosphorous distributions which are outputs of the BiosEquil model by Raupach et al. (2001a; 2001b). These data are available from the Australian Natural Resources Atlas at www.nlwra.gov.au/atlas. Inputs to the pre-European models included meteorological surfaces of daily gridded data at 0.05° spatial resolution (for Australia) (Jeffrey et al. 2001), soil characteristics for current conditions derived from the Atlas of Australian Soils (McKenzie et al. 2000), and vegetation characteristics (Leaf Area Index) (Lu et al. 2001). The 0.05° gridded data were resampled to 0.01° using the cubic algorithm with RESAMPLE in ARCINFO GRID. Zero values were assumed to represent NODATA values (e.g. lakes) and were iteratively filled using the DATA option of the FOCALMEAN command with a CIRCLE expand radius of 3 cells in ARCINFO GRID, as described above.

溶解态磷（dissolved phosphorous）浓度的计算方式为植物可利用矿物磷库与土壤水库的比值。降雨对这两个库的分布具有显著影响，但并不影响二者的比值。溶解态磷浓度的分布可由空气干燥度（饱和水汽压差，saturation deficit）对溶解态磷库的影响加以解释——该影响主要由净初级生产力（Net Primary Productivity，NPP）调控，且强于空气干燥度对土壤水库的影响（土壤水库主要由降雨或能量限制决定）。这一机制使得土壤水中的磷浓度分布呈现如下规律：随着气候平均空气干燥度从温带环境向半干旱热带环境升高，磷浓度逐渐降低。本数据集源自Raupach等人（2001a；2001b）开发的BiosEquil模型。

BiosEquil模型的土壤养分输出结果
养分状况是制约澳大利亚植被系统生产力的关键限制因子之一，但目前仅能通过总养分状况这一由《澳大利亚土壤图集》（McKenzie等人，2000）汇编得到的属性进行粗略表征。因此，我们额外汇编了一套0.05°网格分辨率的前欧洲殖民时期（基准期）碳、氮、磷分布预测数据，该数据为Raupach等人（2001a；2001b）开发的BiosEquil模型的输出结果。此类数据可从澳大利亚自然资源图集获取，网址为www.nlwra.gov.au/atlas。
前欧洲殖民时期模型的输入数据包括：澳大利亚区域0.05°空间分辨率的逐日网格化气象表面数据（Jeffrey等人，2001）、源自《澳大利亚土壤图集》（McKenzie等人，2000）的现有土壤特性数据，以及植被特性数据（叶面积指数（Leaf Area Index，LAI））（Lu等人，2001）。
我们采用ARCINFO GRID中的RESAMPLE工具，以三次卷积算法将0.05°的网格化数据重采样至0.01°分辨率。将零值视为无数据（NODATA）值（例如湖泊区域），并使用ARCINFO GRID中FOCALMEAN命令的DATA选项，以圆形邻域扩展半径3个栅格单元进行迭代填充，具体方法如前文所述。