Vegetation - San Mateo County [ds3021]

In 2018, the Golden Gate National Parks Conservancy (Parks Conservancy) (https://parksconservancy.org), non-profit support partner to the National Park Service (NPS) Golden Gate National Recreation Area (GGNRA), initiated a fine scale vegetation mapping project in Marin County. The GGNRA includes lands in San Francisco and San Mateo counties, and NPS expressed interest in pursuing fine scale vegetation mapping for those lands as well. The Parks Conservancy facilitated multiple meetings with potential project stakeholders and was able to build a consortium of funders to map all of San Mateo County (and NPS lands in San Francisco). The consortium included the San Francisco Public Utilities Commission (SFPUC), Midpeninsula Regional Open Space District (MROSD), Peninsula Open Space Trust (POST), San Mateo City/County Association of Governments, and various County of San Mateo departments including Parks, Agricultural Weights and Measures, Public Works/Flood Control District, Office of Sustainability, and Planning and Building. Over a 3-year period, the project, collectively referred to as the “San Mateo Fine Scale Veg Map”, has produced numerous environmental GIS products including 1-foot contours, orthophotography, and other land cover maps. A 106-class fine-scale vegetation map was completed in April 2022 that details vegetation communities and agricultural land cover types, including forests, grasslands, riparian vegetation, wetlands, and croplands. The environmental data products from the San Mateo Fine Scale Veg Map are foundational and can be used by organizations and government departments for a wide range of purposes, including planning, conservation, and to track changes over time to San Mateo County’s habitats and natural resources. Development of the San Mateo fine-scale vegetation map was managed by the Golden Gate National Parks Conservancy and staffed by personnel from Tukman Geospatial (https://tukmangeospatial.com/), Aerial Information Systems (AIS; http://www.aisgis.com/), and Kass Green and Associates. The fine-scale vegetation map effort included field surveys by a team of trained botanists including Neal Kramer, Brett Hall, Lucy Ferneyhough, Brittany Burnett, Patrick Furtado, and Rosie Frederick. Data from these surveys, combined with older surveys from previous efforts, were analyzed by the California Native Plant Society (CNPS) Vegetation Program (https://www.cnps.org/vegetation), with support from the California Department of Fish and Wildlife Vegetation Classification and Mapping Program (VegCAMP; https://wildlife.ca.gov/Data/VegCAMP) and ecologists with NatureServe (https://www.natureserve.org/) to develop a San Mateo County-specific vegetation classification. For more information on the field sampling and vegetation classification work San Mateo County Fine Scale Vegetation Map Final Report refer to the final report (https://nrm.dfg.ca.gov/FileHandler.ashx?DocumentID=212663) issued by CNPS and corresponding floristic descriptions (https://nrm.dfg.ca.gov/FileHandler.ashx?DocumentID=212666 and https://nrm.dfg.ca.gov/FileHandler.ashx?DocumentID=212667). Existing lidar data, collected in 2017 by San Mateo County was used to support the project. The lidar point cloud, and many of its derivatives, were used extensively during the process of developing the fine-scale vegetation and habitat map. The lidar data was used in conjunction with optical data. Optical data used throughout the project included 6-inch resolution airborne 4-band imagery collected in the summer of 2018, as well as various dates of National Agriculture Imagery Program (NAIP) imagery. Key data sets used in the lifeform and the enhanced lifeform mapping process include high resolution aerial imagery from 2018, the lidar-derived Canopy Height Model (CHM), and several other lidar-derived raster and vector datasets. In addition, a number of forest structure lidar derivatives are used in the machine learning portion of the enhanced lifeform workflow. In 2020, an enhanced lifeform map was produced which serves as the foundation for the much more floristically detailed fine-scale vegetation and habitat map. The lifeform map was developed using expert systems rulesets in Trimble Ecognition®, followed by manual editing. In 2020, Tukman Geospatial staff and partners conducted countywide reconnaissance field work to support fine-scale mapping. Field-collected data were used to train automated machine learning algorithms, which produced a fully automated countywide fine-scale vegetation and habitat map. Throughout 2021, AIS manually edited the fine-scale maps, and Tukman Geospatial and AIS went to the field for validation trips to inform and improve the manual editing process. In early January of 2022, draft maps were distributed and reviewed by San Mateo County’s community of land managers and by the funders of the project. Input from these groups was used to further refine the map. The countywide fine-scale vegetation map and related data products were made public in April 2022. In total, 106 vegetation classes were mapped. During the classification development phase, minimum mapping units (MMUs) were established for the vegetation mapping project. An MMU is the smallest area to be mapped on the ground. For this project, the mapping team chose to map different features at different MMUs. The MMU is 1/4 acre for agricultural, woody riparian, and wetland herbaceous classes; 1/2 acre for woody upland, upland herbaceous, and bare land classes; 1/5 acre for developed feature types; and 400 square feet for water. Accuracy assessment plot data were collected in 2021 and 2022. Accuracy assessment results were compiled and analyzed in the April of 2022. Overall accuracy of the lifeform map is 98 percent. Overall accuracy of the fine-scale vegetation map is 83.5 percent, with an overall ‘fuzzy’ accuracy of 90.8 percent.2018年，作为美国国家公园管理局（National Park Service，简称NPS）金门国家游乐区（Golden Gate National Recreation Area，简称GGNRA）的非营利支持合作伙伴，金门国家公园保护区联合会（Golden Gate National Parks Conservancy，以下简称公园联合会，https://parksconservancy.org）在马林县启动了一项精细尺度植被制图项目。GGNRA涵盖旧金山与圣马特奥县的土地，NPS亦表达了对上述区域开展精细尺度植被制图的兴趣。公园联合会与潜在项目利益相关方多次召开会议，成功组建了资助联盟，负责绘制圣马特奥县全域（以及旧金山境内的NPS土地）的植被图。该联盟成员包括旧金山公共事业委员会（San Francisco Public Utilities Commission，简称SFPUC）、中半岛区域开放空间区（Midpeninsula Regional Open Space District，简称MROSD）、半岛开放空间信托基金（Peninsula Open Space Trust，简称POST）、圣马特奥市县政府协会，以及圣马特奥县多个部门，包括公园管理局、农业度量衡局、公共工程/防洪管理区、可持续发展办公室与规划和建筑局。该项目历时三年，被统称为“圣马特奥精细尺度植被制图项目”，产出了众多环境地理信息系统（Geographic Information System，简称GIS）产品，包括1英尺分辨率等高线、正射影像以及其他土地覆盖图。2022年4月，包含106个类别的精细尺度植被图正式完成，该图详细记录了植被群落与农业土地覆盖类型，涵盖森林、草原、河岸植被、湿地与农田。圣马特奥精细尺度植被制图项目产出的环境数据产品具有基础性价值，可被各类机构与政府部门用于规划、保护，以及追踪圣马特奥县栖息地与自然资源的长期变化。圣马特奥精细尺度植被图的开发工作由金门国家公园保护区联合会管理，执行团队来自图克曼地理空间公司（Tukman Geospatial，https://tukmangeospatial.com/）、航空信息系统公司（Aerial Information Systems，简称AIS；http://www.aisgis.com/）以及Kass Green and Associates。精细尺度植被制图工作包含由一支经过培训的植物学家团队开展的野外调查，该团队成员包括Neal Kramer、Brett Hall、Lucy Ferneyhough、Brittany Burnett、Patrick Furtado与Rosie Frederick。这些野外调查数据与既往项目的旧有调查数据，经加州本土植物学会（California Native Plant Society，简称CNPS）植被项目（https://www.cnps.org/vegetation）分析，并得到了加州鱼类和野生动物部植被分类与制图项目（Vegetation Classification and Mapping Program，简称VegCAMP；https://wildlife.ca.gov/Data/VegCAMP）以及NatureServe（https://www.natureserve.org/）的生态学家支持，最终开发出适用于圣马特奥县的专属植被分类体系。如需了解野外采样与植被分类工作的更多详情，请参阅CNPS发布的《圣马特奥县精细尺度植被制图最终报告》（https://nrm.dfg.ca.gov/FileHandler.ashx?DocumentID=212663）以及对应的植物区系描述文档（https://nrm.dfg.ca.gov/FileHandler.ashx?DocumentID=212666与https://nrm.dfg.ca.gov/FileHandler.ashx?DocumentID=212667）。项目使用了圣马特奥县2017年收集的现有激光雷达（LiDAR）数据。激光雷达点云及其诸多衍生产品在精细尺度植被与栖息地制图的开发过程中得到了广泛应用。该激光雷达数据与光学数据结合使用。项目中使用的光学数据包括2018年夏季采集的6英寸分辨率机载四波段影像，以及不同时期的美国农业部农业影像计划（National Agriculture Imagery Program，简称NAIP）影像。在生命型与增强型生命型制图流程中使用的关键数据集包括2018年的高分辨率航空影像、激光雷达衍生的冠层高度模型（Canopy Height Model，简称CHM），以及其他多个激光雷达衍生的栅格与矢量数据集。此外，增强型生命型工作流的机器学习部分还使用了多种森林结构激光雷达衍生产品。2020年，研究团队产出了增强型生命型图，该图为细节更为丰富的精细尺度植被与栖息地制图奠定了基础。该生命型图基于专家系统规则集在Trimble Ecognition®软件中开发完成，随后进行了人工编辑。2020年，图克曼地理空间公司的员工与合作伙伴开展了全县范围的踏勘野外工作，以支撑精细尺度制图。野外采集的数据被用于训练自动化机器学习算法，进而生成了全县范围的全自动精细尺度植被与栖息地图。2021年全年，AIS对精细尺度地图进行了人工编辑，图克曼地理空间公司与AIS还前往野外开展验证工作，以优化人工编辑流程。2022年1月初，草案地图被分发至圣马特奥县的土地管理者群体与项目资助方进行评审。来自这些群体的意见被用于进一步优化地图。全县范围的精细尺度植被图及相关数据产品于2022年4月正式公开。本次制图共涵盖106个植被类别。在分类体系开发阶段，项目组为植被制图项目设定了最小制图单元（Minimum Mapping Unit，简称MMU）。MMU指地面上可被制图的最小面积。本项目中，制图团队针对不同地物设定了不同的MMU：农业、木本河岸与草本湿地类别的MMU为1/4英亩；木本高地、高地草本与裸地区类别的MMU为1/2英亩；开发地物类别的MMU为1/5英亩；水体类别的MMU为400平方英尺。精度评估样地数据采集于2021年与2022年。精度评估结果于2022年4月汇总分析。生命型图的总体精度为98%。精细尺度植被图的总体精度为83.5%，总体“模糊”精度为90.8%。