ABSTRACT:

Annual peak flows for the CAMELS basins; downloaded from USGS National WaterInformation System (https://nwis.waterdata.usgs.gov/usa/nwis/peak; accessed 10 June 2021)

ABSTRACT:

Peak flow observations for each of the CAMELS basins from the USGS National Water Information System (https://nwis.waterdata.usgs.gov/usa/nwis/peak; accessed 10 June 2021).

We used all available annual peak flows for each of the CAMELS basins and fit these values to the Pearson Type III distribution with log transformation using the method of moments, as described in U.S. Interagency Committee on Water Data,Bulletin 17b (IACWD, 1982). The probability density function is: f(x|\tau, \alpha, \beta) = \frac{(\frac{x-\tau}{\beta})^{\alpha-1}exp(-\frac{x-\tau}{\beta})}{|\beta|\gamma(\alpha)}, The CAMELS basins are suitable for this method under the assumptions of flood flows that are not appreciably altered by reservoir regulation, watershed changes or where the possibility of unusual events, such as dam failures.

We used Matlab code from Mathworks File Exchange to fit the peak annual flow events to the distribution to obtain return period estimates for each basin (Burkey, 2009). These return period calculations can also be done with free and open source software available from the USGS (https://water.usgs.gov/software/PeakFQ/; accessed 10 June 2021). We classified the water year of each basin (basin-year) according to the return period of its observed peak annual discharge

ABSTRACT:

Runs from two papers exploring the use of mass conserving LSTM. Model results used in the papers are 1) model_outputs_for_analysis_extreme_events_paper.tar.gz, and 2) model_outputs_for_analysis_mass_balance_paper.tar.gz.

The models here are trained/calibrated on three different time periods. Standard Time Split (time split 1): test period(1989-1999) is the same period used by previous studies which allows us to confirm that the deep learning models (LSTM andMC-LSTM) trained for this project perform as expected relative to prior work. NWM Time Split (time split 2): The second test period (1995-2014) allows us to benchmark against the NWM-Rv2, which does not provide data prior to 1995. Return period split: The third test period (based on return periods) allows us to benchmark only on water years that contain streamflow events that are larger (per basin) than anything seen in the training data (<= 5-year return periods in training and > 5-year return periods in testing).

Also included are an ensemble of model runs for LSTM, MC-LSTM for the "standard" training period and two forcing products. These files are provided in the format "<model_type>_<forcing_product>_standard_training.tar.gz". Note that these individual ensemble member runs we used to produce the runs in the files "model_outputs_for_analysis_<*>_paper.tar.gz".

IMPORTANT NOTE: This python environment should be used to extract and load the data: https://github.com/jmframe/mclstm_2021_extrapolate/blob/main/python_environment.yml, as the pickle files serialized the data with specific versions of python libraries. Specifically, the pickle serialization was done with xarray=0.16.1.

Code to interpret these runs can be found here:
https://github.com/jmframe/mclstm_2021_extrapolate
https://github.com/jmframe/mclstm_2021_mass_balance

Papers are available here:
https://hess.copernicus.org/preprints/hess-2021-423/

ABSTRACT:

This repository holds supplemental figures for the paper "On Strictly Enforced Mass Conservation Constraints for Modeling the Rainfall-Runoff Process". These include distributions of event-based runoff ratios, scatter plots comparing event-based runoff ratios, initial (antecedent) flows, and rainfall totals, and hydrographs of every qualifying "event" that spans both Daymet and NLDAS forcing data.

ABSTRACT:

This repository holds supplemental figures for the paper "On Strictly Enforced Mass Conservation Constraints for Modeling the Rainfall-Runoff Process". These include distributions of event-based runoff ratios, scatter plots comparing event-based runoff ratios, initial (antecedent) flows, and rainfall totals, and hydrographs of every qualifying "event" that spans both Daymet and NLDAS forcing data.
Contents: 1) "hydrographs" is a directory with images showing hydrographs for every event in the record where the event criteria is met by both NLDAS and Daymet forcings coincidentally, 2) "basin_event_distributions_and_scatter_plots" is a directory with images showing distributions of runoff coefficients (runoff ratios) from 100 nearest-neighbour events; empirical membership values of potential runoff coefficients from nearest-neighbour storms based on Mahalabonis distances based on antecedent streamflow and rainfall total, and 3) "regional_q_q_plots" is a directory with images showing quantile-quantile plots of the percentile bins of the count of the events against the corresponding idealized percentile count. This is commonly referred to as a quantile-quantile plot. We then also plot the distance (and cumulative distance) of these quantiles from the one-to-one line.

ABSTRACT:

This project contains preliminary 1D and 2D hydraulic models for a reach the Arroyo Seco (Los Angeles River tributary) upstream of the Brown Mountain Debris Barrier Dam. These models are **not** ready for use, they are **only** a starting point for further analysis.

ABSTRACT:

This dataset supports the analysis presented in our study on flood prediction using the Novel Water Model-Convolutional Neural Network (NWM-CNN). Aimed at enhancing flood forecasting and resilience, this collection encompasses a comprehensive compilation of flood events across California during the Atmospheric River events during the water year 2023. Included are NWM-CNN predictions, Sentinel-1 flood observations, and historical flood damage reports for Sacramento County, California, during the significant 2023 atmospheric river event and over the past decades.

The dataset is structured to facilitate a detailed examination of the NWM-CNN model's performance in predicting surface water area with high temporal resolution and accuracy. By integrating satellite imagery with hydrodynamic modeling, the NWM-CNN model represents a significant advancement in flood modeling, offering an effective tool for damage assessment, flood forecasting, and supporting parametric insurance solutions.

Key components of the dataset include:
* NWM-CNN Predictions: Model predictions at a 250m grid cell resolution.
* Sentinel-1 Flood Observations: Satellite-derived flood extents used for comparison against model predictions.
* Flood Damage Reports: Historical records of flood damage within Sacramento County, providing a ground-truth comparison for model efficacy.

By making this data publicly available, we aim to contribute to the collective efforts in reducing the impacts of climate disasters through improved access to scientific information and resources.