Artificial Intelligence Deep Learning Model for Mapping Wetlands Yields 94% Accuracy

安纳波利斯, MD -十大赌博正规老平台协会的数据科学团队开发了一个人工智能深度学习模型，用于绘制湿地地图, which resulted in 94% accuracy. Supported by EPRI, an independent, non-profit energy research and development institute; Lincoln Electric System; and the Grayce B. Kerr Fund, Inc., 这种湿地制图方法可以为保护和保存湿地提供重要的成果. The results are published in the peer-reviewed journal Science of the Total Environment.

该团队训练了一个机器学习(卷积神经网络)模型，使用来自三个地区的免费数据进行高分辨率(1m)湿地测绘, Minnesota; Kent County, Delaware; and St. Lawrence County, New York. The full model, 这需要国家湿地数据和国家湿地清单(NWI)提供的当地培训数据。, mapped 湿地 with 94% accuracy.

“我们很高兴支持这个令人兴奋的项目，因为它探索了利用卫星图像划定湿地的新方法,” said EPRI Principal Technical Leader Dr. Nalini饶. “通过在办公桌上使用GIS工具，它有可能节省自然资源管理人员在现场的时间. +, 它可以帮助公司和公众管理对湿地的影响，因为计划建设基础设施以实现脱碳目标.”

“《十大赌博正规老平台》将数千亿美元投入到将对景观产生影响的项目中. 然而, the data that we rely on to minimize impacts to 湿地 is distressingly outdated,环境政策创新中心恢复经济中心主任贝卡·马德森说, a former EPRI researcher. “现在是更新我们国家湿地数据的最佳时机，并建立一个可持续的、具有成本效益的过程来保持它们的更新.”

“当这个高度精确的模型被放大到更大的地理区域，比如切萨皮克湾或美国本土的湿地时, this will be a game changer. 它避免了人工绘制湿地地图的需要，也避免了传统的机器学习绘制湿地地图需要大量的数据处理, curation and manual feature engineering, both of which are time-consuming, labor intensive and very expensive,” said Chesapeake Conservancy’s Data Science Lead/Senior Data Scientist Dr. Kumar Mainali.

What This Means for Protecting & Conserving Wetlands

The new model will help infrastructure planners avoid 湿地 in the planning process, resulting in cost savings and 湿地 conservation. 潜在的有利情况包括不断努力扩大和发展可再生能源, which requires expanding electric power infrastructure.

The product of the model is a map of wetland probability. This probability data may be used to map the most likely wetland extent, but if users prefer, they can map wetland extent with a lower probability threshold. 最终的地图限制了湿地遗漏的可能性，即使它绘制的湿地比现实中存在的更多.

也有可能使用这个模型来绘制湿地已经消失的位置，因为它们是用NWI绘制的. Additionally, potential locations for wetland restoration could also be identified. 例如, 持续潮湿的农田被模型所捕捉，即使是为了野外湿地的描绘, these areas are not considered 湿地 when actively farmed.

下一个步骤

该团队将把模型扩展到各州或更大的地区，并继续在不同的地理位置上训练模型.

Model Overcomes Outdated Data in 内布拉斯加州 Pilot

Following the initial model development, the model was extended to include Lancaster County, 内布拉斯加州. 在该地区建立湿地模型具有挑战性，因为该地区的NWI数据已经过时了几十年, and included 湿地 in several areas where they had been lost to development. 该团队有兴趣了解该模型是否可以成功地绘制湿地地图，因为最近没有高质量的湿地数据集来训练该模型.

湿地模型是用几十年前的NWI数据集和最近的卫星和航空图像数据进行训练的. 研究小组发现，与训练前的预测相比，NWI数据将湿地测绘的局部精度提高了10%, showing the importance of using local training data in new geographies. 除了, the model correctly omitted 湿地 where they had been lost to development, despite these 湿地 remaining in the outdated training data, as shown in the image below (outdated training data shown in green; model prediction in purple, overlaid over recent satellite imagery). 该模型在确定数据中的主导模式方面的表现既提高了局部制图精度，又能准确地反映湿地的存在和不存在，这对该方法的实用性有很大的帮助.

尽管湿地数据在规划基础设施项目和管理野生动物方面发挥着重要作用, NWI 湿地 data have not been comprehensively updated for many years. As shown in the map below, much NWI data across the nation dates to the 1970s and 1980s, yet remains the best available data. 一种可以利用不同年份的训练数据的湿地制图建模方法将在最需要的地方实现湿地制图的现代化，这将是非常有用的. (For more information, see “It’s Time to Invest in a Modern Map of Our Nation’s Wetlands”: http://www.policyinnovation.org/blog/investing-in-a-modern-map-of-our-nations-湿地)

 

About the Model

在湿地训练中使用的“预测”层，模型从中学习湿地中发现的模式:美国农业部国家农业图像计划(NAIP)航空图像(1m), Sentinel-2 optical satellite imagery (10-20m), LiDAR-derived geomorphons, an approach to mapping landforms that Chesapeake Conservancy has been applying to advance high-resolution stream mapping; and LiDAR intensity, an index that is frequently used to identify water and persistently wet soils.

Additionally, 该团队仅使用USDA NAIP和Sentinel-2数据作为输入层，训练了一个更简单的模型, securing an accuracy of 91.6%.

该论文的共同作者是十大赌博正规老平台协会的数据科学主管/高级数据科学家Kumar Mainali, Ph.D., Senior Data Scientist Michael Evans, Ph.D., Geospatial Technology Manager Emily Mills (formerly with Chesapeake Conservancy), Senior Geospatial Technical Lead David Saavedra, Vice President of Climate Strategy Susan Minnemeyer, and former EPRI Project Manager Becca Madsen, now with the Environmental Policy Innovation Center.

Read the publication “基于开放数据的高分辨率湿地映射的卷积神经网络:变量选择和可推广模型的挑战在线 Science of the Total Environment.

For more information, see digital StoryMap, “Identifying Wetlands with Deep Learning: How an EPRI & 十大赌博正规老平台协会合作改进桌面湿地识别以改进规划.”

###