Parcel Manager Data set

Data contained in this repository can be used for replicating the two use cases of Parcel Manager software application presented in the paper (Colomb et al., 2021, submitted).

The first use case enables the simulation of a scenario called test scenario designed specifically to test different parcel division processes and workflows with Parcel Manager. The area under study is a small community (Gennes, 681 inhabitants) located in the east of France. Data used are from the French IGN BD Topo 2018.

The second use case enables the comparison of the shape of parcels created by simulation with the shape of parcels in real cases. The comparison concerns the parcel plans in 2003 and 2018 of 11 communities of the Seine-et-Marne department, near Paris capital city (France). Data for 2018 are from the French IGN BD Topo 2018. They have been cleaned manually as follows:
– removing parcels that represent driveways or roads which could be used by residential parcels,
– removing the tiny parcels resulting from a former division process and that could prevent the access to roads of neighbouring parcels,
– removing parcels located on water surfaces or railways,
– removing roads of type ’trails’ and ’stairs’.

Parcel data for 2003 are from the IGN BD-Parcellaire 2003. Building and road data come from the IGN BD-Topo 2005. The cleaning results of the 2018 parcel plan have been copied into the 2003 parcel plan.

The road network encompasses every types of paths that are accessible with a car.
– Driveways are small paths for cars that connect a house to the road network. Driveways are included in parcels and are not considered as a proper road.
– Gravel roads are non-asphalted trails. They can occasionally be used by cars. Their level of attraction is low (level 2).
– Lanes are small roads dedicated to house access. Their level of attraction is maximal (level 4). Note that peripheral roads created with the straight skeleton algorithm are classified as lanes.
– Streets are roads that connect lanes. Their level of attraction is high (level 3).
– Arterial roads are high-speed roads connecting communities. Their level of attraction for parcel contact is minimal (level 1).

In the data set created for the test Scenario, the road layer has been manually enriched: missing trails and lanes have been added; type and level of attraction of some road segments have been changed.

本仓库所包含的数据可用于复现该论文（Colomb等人，2021年，已投稿）中提出的两款地块管理器（Parcel Manager）软件应用场景。

首个使用场景可模拟名为测试场景（test scenario）的工况，该场景专为使用地块管理器（Parcel Manager）测试不同地块划分流程与工作流而设计。研究区域为法国东部的一座小型社区（热讷斯，Gennes，人口681），所用数据源自法国国家地理信息研究院（IGN）2018年版BD Topo数据集。

第二个使用场景可将模拟生成的地块形状与真实地块形状进行对比。对比对象为法国首都巴黎附近塞纳-马恩省（Seine-et-Marne）11个社区2003年与2018年的地块规划图。2018年的数据源自法国国家地理信息研究院（IGN）2018年版BD Topo数据集，已通过如下方式完成人工清洗：
1. 移除可作为住宅地块通行路径的车道或道路；
2. 移除由既往划分流程产生的微型地块，此类地块可能会阻碍相邻地块的道路通行；
3. 移除位于水面或铁路沿线的地块；
4. 移除“步道”与“阶梯”类道路。

2003年的地块数据源自法国国家地理信息研究院（IGN）2003年版BD-Parcellaire数据集；建筑与道路数据源自法国国家地理信息研究院（IGN）2005年版BD Topo数据集。研究人员已将2018年地块规划图的清洗结果同步至2003年的地块规划图中。

本数据集涵盖所有可供汽车通行的道路类型，具体分类如下：
- 车道（Driveways）：连接住宅与道路网络的小型汽车通行路径，属于地块内部设施，不视为正式道路；
- 碎石路（Gravel roads）：未铺设沥青的步道，偶尔可供汽车通行，通行吸引力等级为2（较低）；
- 专用车道（Lanes）：专供住宅通行的小型道路，通行吸引力等级为4（最高）。需注意，通过直骨架算法（straight skeleton algorithm）生成的外围道路将被归类为专用车道；
- 街道（Streets）：连接专用车道的道路，通行吸引力等级为3（较高）；
- 主干道（Arterial roads）：连接各社区的高速道路，与地块接触的通行吸引力等级为1（最低）。

针对本次测试场景构建的数据集，已通过人工方式丰富了道路图层：补充了缺失的步道与专用车道，并调整了部分路段的道路类型与通行吸引力等级。