Peter Regier

Peter Regier

Auto-generated profile, based on publicly available information

Peter Regier

Marine and Coastal Research Laboratory Pacific Northwest National Laboratory Sequim WA USA

Research Focus & Metrics

Academic Career Length

11Years (2016-Present)

(Calculated once Scilit publication data is synced: current year − year of first publication + 1)

Representative Publications

View All Publications
30 April 2021 in Nature Communications
Verified
Journal Article
Open access
PDF file
GBGrady Ball
Grady Ball
PRPeter Regier
Peter Regier
RGRicardo González-Pinzón
Ricardo González-Pinzón
JRJustin Reale
Justin Reale
DHDavid Van Horn
David Van Horn
Wildfires are increasing globally in frequency, severity, and extent, but their impact on fluvial networks, and the resources they provide, remains unclear. We combine remote sensing of burn perimeter and severity, in-situ water quality monitoring, and longitudinal modeling to create the first large-scale, long-term estimates of stream+river length impacted by wildfire for the western US. We find that wildfires directly impact ~6% of the total stream+river length between 1984 and 2014, increasing at a rate of 342 km/year. When longitudinal propagation of water quality impacts is included, we estimate that wildfires affect ~11% of the total stream+river length. Our results indicate that wildfire activity is one of the largest drivers of aquatic impairment, though it is not routinely reported by regulatory agencies, as wildfire impacts on fluvial networks remain unconstrained. We identify key actions to address this knowledge gap and better understand the growing threat to fluvial networks, water security, and public health risks.

1 January 2019 in MethodsX
Verified
Journal Article
Open access
PRPeter Regier
Peter Regier
HBHenry Briceño
Henry Briceño
JBJoseph N. Boyer
Joseph N. Boyer
Cumulative sums (Cusums) are a simple, efficient statistical method developed for process control and increasingly used to determine underlying features of time series. Here, two useful applications of Cusums to environmental time series are presented: Cusums in the time domain and plotting Cusum-transformed variables against non-transformed variables to extract meaning in the context of driver-response relationships. These statistical analyses are simple to conduct and provide valuable information about trends, patterns and thresholds of time-series over time and in relation to potential driver variables. In addition, this work investigates the robustness of the Cusum transform to various characteristics of environmental time series that challenge conventional statistical methods. In summary, this work presents: •Cusum methods to derive meaning from complex environmental time series.•Effects of common time series issues on the Cusums method.•Application to real-world datasets.

1 August 2020 in Science of The Total Environment
Verified
Journal Article
Closed access
PRPeter J. Regier
Peter J. Regier
RGRicardo González-Pinzón
Ricardo González-Pinzón
DHDavid J. Van Horn
David J. Van Horn
JRJustin K. Reale
Justin K. Reale
JNJustin Nichols
Justin Nichols
AKAashish Khandewal
Aashish Khandewal

24 February 2023 in Environmental Monitoring and Assessment
Verified
Journal Article
Open access
PDF file
AHA. M. Hopple
A. M. Hopple
KDKennedy O. Doro
Kennedy O. Doro
VBV. L. Bailey
V. L. Bailey
BBB. Bond-Lamberty
B. Bond-Lamberty
NMN. McDowell
N. McDowell
JMJ. Patrick Megonigal
J. Patrick Megonigal
Coastal upland forests are facing widespread mortality as sea-level rise accelerates and precipitation and storm regimes change. The loss of coastal forests has significant implications for the coastal carbon cycle; yet, predicting mortality likelihood is difficult due to our limited understanding of disturbance impacts on coastal forests. The manipulative, ecosystem-scale Terrestrial Ecosystem Manipulation to Probe the Effects of Storm Treatments (TEMPEST) experiment addresses the potential for freshwater and estuarine-water disturbance events to alter tree function, species composition, and ecosystem processes in a deciduous coastal forest in MD, USA. The experiment uses a large-unit (2000 m2), un-replicated experimental design, with three 50 m × 40 m plots serving as control, freshwater, and estuarine-water treatments. Transient saturation (5 h) of the entire soil rooting zone (0–30 cm) across a 2000 m2coastal forest was attained by delivering 300 m3of water through a spatially distributed irrigation network at a rate just above the soil infiltration rate. Our water delivery approach also elevated the water table (typically ~ 2 m belowground) and achieved extensive, low-level inundation (~ 8 cm standing water). A TEMPEST simulation approximated a 15-cm rainfall event and based on historic records, was of comparable intensity to a 10-year storm for the area. This characterization was supported by showing that Hurricane Ida’s (~ 5 cm rainfall) hydrologic impacts were shorter (40% lower duration) and less expansive (80% less coverage) than those generated through experimental manipulation. Future work will apply TEMPEST treatments to evaluate coastal forest resilience to changing hydrologic disturbance regimes and identify conditions that initiate ecosystem state transitions.

11 November 2022 in Limnology and Oceanography: Methods
Verified
Journal Article
Open access
PRPeter Regier
Peter Regier
MDMatthew Duggan
Matthew Duggan
AMAllison Myers‐Pigg
Allison Myers‐Pigg
NWNicholas Ward
Nicholas Ward
Random forests (RF) are an increasingly popular machine learning approach used to model biogeochemical processes in the Earth system. While RF models are robust to many assumptions that complicate deterministic models, there are several important parameterization decisions for appropriate use and optimal model fit. We explored the role that parameter decisions, including training/testing data splitting strategies, variable selection, and hyperparameters play on RF goodness‐of‐fit by constructing models using 1296 unique parameter combinations to predict concentrations of nitrate, a key nutrient for biogeochemical cycling in aquatic ecosystems. Models were built on long‐term, publicly available water quality and meteorology time series collected by the National Estuarine Research Reserve monitoring network for two contrasting ecosystems representing freshwater and brackish estuaries. We found that accounting for temporal dependence when splitting data into training and testing subsets was key for avoiding over‐estimation of model predictive power. In addition, variable selection, the ratio of training to testing data, and to a lesser degree, variables per split and number of trees, were significant parameters for optimizing RF goodness‐of‐fit. We also explored how model parameter decisions influenced interpretation of the relative importance of predictors to the model, and model predictor‐dependent variable relationships, with results suggesting that both data structure and model parameterization influence these factors. Because much of the current RF literature is written for the computational and statistical science communities, the primary goal of this study is to provide guidelines for aquatic scientists new to machine learning to apply RF techniques appropriately to aquatic biogeochemical datasets.

8 August 2024 in Environmental Science & Technology
Verified
Journal Article
Closed access
FMFausto Machado-Silva
Fausto Machado-Silva
MWMichael N. Weintraub
Michael N. Weintraub
NWNicholas David Ward
Nicholas David Ward
KDKennedy O. Doro
Kennedy O. Doro
PRPeter J. Regier
Peter J. Regier
VBVanessa L Bailey
Vanessa L Bailey