Machine Learning Specialist
Office of the Chief Operating Officer
Email:
shamshaw@usgs.gov
Office Phone:
802-324-6221
ORCID:
0000-0002-0583-4237
Location
425 Jordan Road
Troy
, NY
12180
US
Supervisor:
Joel D Blomquist
|
This dataset consists of daily streamflow percentiles for 1981-10-01 to 2020-03-31 relevant to streamflow drought defined using two approaches: Percentiles accounting for flow seasonality (variable threshold percentiles) and those based on the full record of data for each site regardless of season (fixed threshold percentiles). Because of the size of this dataset (99,530,836 rows), it could not be provided as a .csv file, and is instead provided as a .parquet file. Instructions on reading this file using the R programming language are provided in the Processing Step section of this metadata. The daily streamflow percentiles were estimated for ungaged areas of the Colorado River Basin (CRB) using neural network models,...
|
This metadata record describes model outputs and supporting model code for the Data-Driven Drought Prediction project of the Water Resources Mission Area Drought Program. The data listed here include outputs of multiple machine learning model types for predicting hydrological drought at select locations within the conterminous United States. The child items referenced below correspond to different models and spatial extents (Colorado River Basin region or conterminous United States). See the list below or metadata files in each sub-folder for more details. Daily streamflow percentile predictions for the Colorado River Basin region — Outputs from long short-term memory (LSTM) deep learning models corresponding to...
Tags: Conterminous United States,
Hydrology,
USGS Science Data Catalog (SDC),
Upper Colorado River,
climatologyMeteorologyAtmosphere, All tags...
deep learning,
drought prediction,
droughts,
hydrology,
long short-term memory,
machine learning,
modeling,
river systems,
statistical analysis,
streamflow,
streamflow percentiles,
streamflow predictions, Fewer tags
|
This metadata record describes outputs from 12 configurations of long short-term memory (LSTM) models which were used to predict streamflow drought occurrence at 384 stream gage locations in the Colorado River Basin region. The models were trained on data from 01-Oct-1981 to 31-Mar-2005 and validated over the period of record spanning 01-Apr-2005 to 31-Mar- 2014. The models use explanatory variable inputs described in Wieczorek (2023) (doi.org/10.5066/P98IG8LO) to predict daily streamflow and streamflow percentiles as described in Simeone (2022) (doi.org/10.5066/P92FAASD). Separate models were trained to predict daily streamflow and streamflow percentiles. Two types of percentiles were modeled: (1) fixed-threshold...
Categories: Data;
Tags: Colorado River Basin,
climatologyMeteorologyAtmosphere,
deep learning,
drought prediction,
droughts, All tags...
hydrology,
long short-term memory,
machine learning,
modeling,
river systems,
statistical analysis,
streamflow,
streamflow percentiles,
streamflow predictions, Fewer tags
|
|