MA Stream Temperature and Thermal Habitat - Baseline (By thermal class)

This dataset can be further explorer using the MA Stream Temperature and Thermal Habitat Explorer hosted by author Jeffery D Walker, PHD. Study Objectives The primary objective of this study was to generate projections of changes in stream temperature and thermal habitat (i.e., cold water fish habitat) due to climate change across the state of Massachusetts. To achieve this, statistical and machine learning models were developed for predicting stream temperatures based on air temperature and various landscape metrics (e.g., land use, elevation, drainage area). The model was then used in conjunction with climate change projections of air temperature increases to estimate the potential changes in stream temperatures and thermal habitat across the state. The results of this study are made available through this web-based tool to inform conservation and management decisions related to the protection of coldwater fish habitat in Massachusetts Modeling Methodology A regional model was developed for predicting stream temperatures in all streams and rivers across the state, excluding the largest rivers such as the Connecticut and Merrimack. The model was comprised of two components: 1) a non-linear regression model representing the functional relationship between air and water temperatures at a single location, and 2) a machine learning model (boosted decision trees) for estimating the parameters of the air-water temperature model spatially based on landscape characteristics. Together, these models demonstrated strong performance in predicting weekly water temperatures with an RMSE of 1.3 degC and Nash Sutcliffe Efficiency (NSE) of 0.97 based on an independent subset of the observed data that was excluded from model development and training. Results Under historical baseline conditions (average air temperatures over 1971-2000), the model results showed more abundant cold water habitat in the western part of the state compared to the eastern and coastal areas. Forest and tree canopy cover were among the most important predictors of the relationship between air and water temperatures. The amount of impounded water due to dams upstream of each reach was also important. The majority of cold water habitat (82% of all river miles) were found in first order streams (i.e., headwaters), which are also the most abundant accounting for 60% of all river miles overall. The Deerfield and Hudson-Hoosic drainage basins had the most cold water habitat, which accounted for 80% or more of the total river miles within each basin. Coastal basins such as Narragansett, Piscataqua-Salmon Falls, Charles River, and Cape Code each had less than 5% cold water habitat. Using a series of projected air temperature increases for the RCP 8.5 emissions scenario, the model predicted a reduction in cold water habitat (mean July temp < 18.45 °C) from 30% to 8.5% (a 72% reduction) statewide by the 2090 averaging period (2080-2100). Furthermore, projections for larger streams (orders 3–5) were projected to shift from predominately cool-water (18.45–22.30 °C) to the majority (> 50%) of river miles being classified as warm-water habitat (> 22.30 °C). Conclusions The projected stream temperatures and thermal classifications generated by this project will be a valuable dataset for researchers and resource managers to assess potential climate change impacts on thermal habitats across the state. With this spatially continuous dataset, researchers and managers can identify specific reaches or basins projected to be the most resilient to climate change, and prioritize them for protection or restoration. As more datasets become available, this model can be readily extended and adapted to increase its spatial extent and resolution, and to incorporate flow data for assessing the impacts of not only rising air temperatures but also changing precipitation patterns. Acknowledgements Thanks to Jenn Fair (USGS) for her technical review of the model report and assistance in data gathering at the beginning of the project. I would also like to thank Ben Letcher (USGS) for his feedback and long-term collaboration on EcoSHEDS, which led to this project; Matt Fuller (USDA FS), Jenny Rogers (UMass Amherst), Valerie Ouellet (NOAA NMFS), and Aimee Fullerton (NOAA NMFS) for taking the time to discuss their experience, insights, and ideas regarding regional stream temperature modeling; Lisa Kumpf (CRWA) and Ryan O’Donnell (IRWA) for sharing their data directly; and Sean McCanty (NRWA), Julia Blatt (Mass Rivers Alliance), and Sarah Bower (Mass Rivers Alliance) for their assistance in sending out a request to the Mass Rivers Alliance for stream temperature data. Lastly, I am grateful for the countless individuals who collected the temperature data and without whom this project would not have been possible. Funding This study was performed by Jeffrey D Walker, PhD of Walker Environmental Research LLC in collaboration with MA Division of Fisheries and Wildlife (MassWildlife). Funding was provided by the 2018 State Hazard Mitigation and Climate Adaptation Plan (SHMCAP) for Massachusetts.本数据集可通过由杰弗里·D·沃克博士（Jeffery D Walker, PHD）主持开发的MA溪流温度与热栖息地探索器进行进一步探索。 研究目标 本研究的核心目标为：针对马萨诸塞州全域，预测气候变化引发的溪流温度与热栖息地（即冷水鱼类栖息地）变化情况。为达成该目标，本研究构建了统计与机器学习模型，基于气温及各类景观指标（如土地利用、海拔、流域面积）预测溪流温度。随后，将该模型与气温升高的气候变化预测结果结合，估算全州范围内溪流温度与热栖息地的潜在变化。本研究成果通过该基于网页的工具对外发布，可为马萨诸塞州冷水鱼类栖息地保护相关的保护与管理决策提供参考依据。 建模方法 本研究构建了一套区域模型，用于预测全州除康涅狄格河、梅里马克河等大型河流外的所有溪流与河道的温度。该模型包含两个组成部分：1） 用于表征单点位置气温与水温之间函数关系的非线性回归模型；2） 基于景观特征，对气温-水温模型的参数进行空间估算的机器学习模型（提升决策树（boosted decision trees））。基于从模型开发与训练阶段排除的独立观测数据子集，上述联合模型在预测周均水温时表现优异，均方根误差（Root Mean Square Error, RMSE）为1.3 ℃，纳什效率系数（Nash Sutcliffe Efficiency, NSE）达0.97。 研究结果 在历史基准情景下（1971-2000年平均气温），模型结果显示，全州西部的冷水栖息地资源较东部与沿海地区更为丰富。森林与林冠覆盖度是影响气温-水温关系的最重要预测因子之一，各河段上游大坝形成的蓄水体量同样具有重要影响。绝大多数冷水栖息地（占河道总里程的82%）分布于一级溪流（first order streams，即源头溪流（headwaters））中，而一级溪流本身也占全州河道总里程的60%。迪尔菲尔德流域与哈德逊-胡西克流域拥有最多的冷水栖息地，占各自流域内河道总里程的80%及以上。纳拉甘西特（Narragansett）、皮斯卡塔夸-鲑鱼瀑布（Piscataqua-Salmon Falls）、查尔斯河（Charles River）及科德角（Cape Code）等沿海流域的冷水栖息地占比均不足5%。 针对典型浓度路径8.5（Representative Concentration Pathway 8.5, RCP 8.5）排放情景下的一系列气温升高预测，模型预测到2090年平均时段（2080-2100年），全州冷水栖息地（7月平均水温<18.45 ℃）占比将从30%降至8.5%（降幅达72%）。此外，预测结果显示，较大级别的溪流（3-5级）将从以凉水生境（18.45~22.30 ℃）为主，转变为大部分（>50%）河道里程被归类为暖水生境（>22.30 ℃）。 结论 本项目生成的预测溪流温度与热栖息地分类结果，将成为研究人员与资源管理者评估气候变化对全州热栖息地潜在影响的宝贵数据集。借助这套空间连续型数据集，研究人员与管理者可识别出被预测为对气候变化适应性最强的特定河段或流域，并优先开展保护与修复工作。随着更多数据集的公开，该模型可轻松进行扩展与适配，以扩大其空间覆盖范围与分辨率，并纳入径流数据，从而不仅评估气温升高的影响，还可分析降水模式变化带来的效应。 致谢 感谢珍妮·费尔（Jenn Fair，美国地质调查局（United States Geological Survey, USGS））对模型报告的技术审阅，以及项目初期在数据收集方面提供的协助。同时感谢本·莱彻（Ben Letcher，USGS）的反馈意见，以及其在EcoSHEDS项目上的长期合作，正是这些促成了本研究的开展；感谢马特·富勒（Matt Fuller，美国农业部林务局（United States Department of Agriculture Forest Service, USDA FS））、珍妮·罗杰斯（Jenny Rogers，马萨诸塞大学阿默斯特分校）、瓦莱丽·乌莱特（Valerie Ouellet，美国国家海洋和大气管理局国家海洋渔业局（National Oceanic and Atmospheric Administration National Marine Fisheries Service, NOAA NMFS））及艾米·富勒顿（Aimee Fullerton，NOAA NMFS）抽出时间，分享他们在区域溪流温度建模方面的经验、见解与想法；感谢丽莎·坎普夫（Lisa Kumpf，康涅狄格河流域联盟（CRWA））与瑞安·奥唐奈（Ryan O’Donnell，伊普斯维奇河流域联盟（IRWA））直接分享他们的数据；感谢肖恩·麦坎蒂（Sean McCanty，NRWA）、朱莉娅·布拉特（Julia Blatt，马萨诸塞河流联盟）及萨拉·鲍尔（Sarah Bower，马萨诸塞河流联盟）协助向马萨诸塞河流联盟发起溪流温度数据征集请求。最后，衷心感谢所有参与温度数据采集的人员，若无他们的工作，本研究将无法完成。 资助说明 本研究由沃克环境研究有限责任公司的杰弗里·D·沃克博士（Jeffrey D Walker, PhD）与马萨诸塞州鱼类与野生动物管理局（MassWildlife）合作开展。资助资金来自马萨诸塞州2018年州灾害减缓与气候适应计划（State Hazard Mitigation and Climate Adaptation Plan, SHMCAP）。