Centre for Ecology and Hydrology Edinburgh, Bush Estate, Penicuik, Midlothian EH26 0QB, UK; School of GeoSciences, The University of Edinburgh, Drummond Street, Edinburgh EH8 9XP, UK; IVL Swedish Environmental Research Institute Ltd, P.O. Box 5302, SE-400 14 Gothenburg, Sweden.
Centre for Ecology and Hydrology Edinburgh, Bush Estate, Penicuik, Midlothian EH26 0QB, UK; School of GeoSciences, The University of Edinburgh, Drummond Street, Edinburgh EH8 9XP, UK.
Environ Pollut. 2018 Sep;240:412-421. doi: 10.1016/j.envpol.2018.04.132. Epub 2018 May 9.
Ammonia emissions vary greatly at a local scale, and effects (eutrophication, acidification) occur primarily close to sources. Therefore it is important that spatially distributed emission estimates are located as accurately as possible. The main source of ammonia emissions is agriculture, and therefore agricultural survey statistics are the most important input data to an ammonia emission inventory alongside per activity estimates of emission potential. In the UK, agricultural statistics are collected at farm level, but are aggregated to parish level, NUTS-3 level or regular grid resolution for distribution to users. In this study, the Modifiable Areal Unit Problem (MAUP), associated with such amalgamation, is investigated in the context of assessing the spatial distribution of ammonia sources for emission inventories. England was used as a test area to study the effects of the MAUP. Agricultural survey data at farm level (point data) were obtained under license and amalgamated to different areal units or zones: regular 1-km, 5-km, 10-km grids and parish level, before they were imported into the emission model. The results of using the survey data at different levels of amalgamation were assessed to estimate the effects of the MAUP on the spatial inventory. The analysis showed that the size and shape of aggregation zones applied to the farm-level agricultural statistics strongly affect the location of the emissions estimated by the model. If the zones are too small, this may result in false emission "hot spots", i.e., artificially high emission values that are in reality not confined to the zone to which they are allocated. Conversely, if the zones are too large, detail may be lost and emissions smoothed out, which may give a false impression of the spatial patterns and magnitude of emissions in those zones. The results of the study indicate that the MAUP has a significant effect on the location and local magnitude of emissions in spatial inventories where amalgamated, zonal data are used.
氨气排放具有显著的局地变异性，其影响（富营养化、酸化）主要发生在污染源附近。因此，准确确定空间分布的排放估算位置非常重要。氨气排放的主要来源是农业，因此农业调查统计数据是氨排放清单的最重要输入数据之一，同时还需要根据活动排放潜力进行估算。在英国，农业统计数据是在农场层面收集的，但为了分发给用户，这些数据会被汇总到教区、NUTS-3 级别或常规网格分辨率层面。在本研究中，在评估排放清单中氨气源的空间分布时，研究了与这种合并相关的可修改面积单位问题（MAUP）。英格兰被用作试验区，以研究 MAUP 的影响。根据许可，在将农场层面（点数据）的农业调查数据合并到不同的面积单位或区域之前，首先将其合并到常规的 1 公里、5 公里和 10 公里网格以及教区层面：在将其导入排放模型之前。评估了使用不同合并水平的调查数据的结果，以估计 MAUP 对空间清单的影响。分析表明，应用于农场层面农业统计数据的聚合区的大小和形状强烈影响模型估算的排放位置。如果区域太小，可能会导致错误的排放“热点”，即实际上并不局限于分配给它们的区域的人为高排放值。相反，如果区域太大，则可能会丢失细节并使排放平滑化，从而可能对这些区域的排放空间模式和幅度产生错误的印象。研究结果表明，在使用合并的、区域数据的空间清单中，MAUP 对排放的位置和局部幅度有重大影响。
