CONUS NG-IDF Data Sets

The NG-IDF datasets, spanning 1951‒2013, characterize the magnitude, trend, seasonality, and driving mechanism of extreme events relevant to hydrologic design for over 200,000 locations across the CONUS at a 1/16th-degree resolution. These datasets enhance traditional precipitation-based intensity-duration-frequency curves (PREC-IDF) by accounting for total water reaching the land surface from rainfall, snowmelt, and rain-on-snow (ROS), i.e., Next-Generation intensity-duration-frequency analysis (NG-IDF).
PREC-IDF's neglect of snow processes can lead to significant biases in extreme event analysis, especially in snowy regions. NG-IDF addresses this by including snow processes, providing a systematic and consistent technique for all environments from rain-dominated, transitional, to snow-dominated locations. The resulting datasets are spatially continuous, and readily usable for supporting site-specific infrastructure design and for assessing the potential biases and design risks related to the use of PREC-IDF.

Abbreviations:

P = Precipitation
R = Rainfall
ROS = Rain-on-snow
M = Snowmelt 
W = Water reaching the land surface from rain, snowmelt, and ROS events

Classification of Driving Mechanism for Extreme W Events:
For each location, the driving mechanism was determined for extreme W events with different durations and return periods, following the classification criteria as follows:

Rainfall (R): precipitation on snow-free ground; 
Snowmelt (M): decreasing SWE, daily rainfall < 10 mm, and the sum of rainfall and snowmelt has < 20% contribution from rainfall.
Rain-on-snow (ROS): decreasing SWE with at least 10 mm of daily rainfall falling on a snowpack with at least 10 mm daily SWE, and snowmelt contributes at least 20% of the total of rain and snowmelt

The datasets include:
Note: In data format description, C = column, R = row, C[i] indicates the ith column of a data file. All data files are prepared in comma-separated value (.csv) format.

list.csv

Description: The geographic coordinates and cluster ID (used for snow parameterization) of 207,173 locations over the CONUS
Data Dimension: 207,173 (R) x 3 (C)
Format: C1: Latitude; C2: Longitude; C3: cluster ID (ranging from 1‒5)

AMF_WY/

Description: Annual maximum series with durations of 24h, 48h, and 72h, driven by different hydrometeorological mechanisms over water years 1951‒2013 (10/1/1950-09/30/2013). The mechanisms include W, P, R, ROS, and M.
Subfolder/File Naming Convention: [duration]/[mechanism].csv, e.g., 24h/W.csv
Data Dimension: 207,173 (R) x 65 (C)
Unit: mm
Format: C1: Latitude; C2: Longitude; C3‒C65: maximum value for each year from 1951-2013

AMF_CY/

Description: Annual maximum series with durations of 24h, 48h, and 72h, driven by different hydrometeorological mechanisms over calendar years 1950‒2012 (1/1/1950-12/30/2012). The mechanisms include W, P, R, ROS, and M.
Subfolder/File Naming Convention: [duration]/[mechanism].csv, e.g., 24h/W.csv
Data Dimension: 207,173 (R) x 65 (C)
Unit: mm
Format: C1: Latitude; C2: Longitude; C3‒C65: maximum value for each year from 1950-2012

IDF/

Description: Discrete IDF values, i.e., the magnitude of extreme events with durations of 24h, 48h, and 72h, driven by different hydrometeorological mechanisms. The mechanisms include W, P, R, ROS, and M.
Subfolder/File Naming Convention: [duration]/[mechanism].csv, e.g., 24h/W.csv
Data Dimension: 207,173 (R) x 9 (C)
Unit: mm
Format: C1: Latitude; C2: Longitude; C3‒C9: IDF values for the return period of 2, 5, 10, 25, 50, 100, and 500 years. NaN indicates no runoff caused by a given mechanism, such as ROS.

IDF_90CI/

Description: 90% C.I. for IDF values in the IDF/ folder described above
Subfolder/File Naming Convention: [duration]/[mechanism]_H.csv, e.g., 24h/W_H.csv

Data Dimension: 207,173 (R) x 9 (C)
Unit: mm
Format: C1: Latitude; C2: Longitude; C3‒C9: 95% quantile for IDF values with the return period of 2, 5, 10, 25, 50, 100, and 500 years. NaN indicates no runoff event caused by a given mechanism.
Subfolder/File Naming Convention: [duration]/[mechanism]_L.csv, e.g., 24h/W_L.csv

Data Dimension: 207,173 (R) x 9 (C)
Unit: mm
Format: C1: Latitude; C2: Longitude; C3‒C9: 5% quantile for IDF values with the return period of 2, 5, 10, 25, 50, 100, and 500 years. NaN indicates no runoff event caused by a given mechanism.

trend/

Description: Sen's slope of Mann-Kendall trend in annual maximum series driven by different hydrometeorological mechanisms over water years 1951‒2013. The mechanisms include W, P, rainfall (R), ROS, and snowmelt (M).
Subfolder/File Naming Convention: [duration]/[mechanism].csv, e.g., 24h/W.csv
Data Dimension: 207,173 (R) x 3 (C)
Unit: mm/year
Format: C1: Latitude; C2: Longitude; C3: Trend indicated by Sen's slope. The value is zero if the trend is not statistically significant.

Driver/

Description: Dominant driving mechanism of extreme W events with different durations and return periods
Subfolder/File Naming Convention: [duration]/[return period].csv, e.g., 24h/50y.csv
Data Dimension: 207,173 (R) x 3 (C)
Format: C1: Latitude; C2: Longitude; C3: Dominant driver IDs (1=R, 2=ROS, 3=M).

risk/

Description: Design risk associated with PREC-IDF estimated 100-year extreme events
Subfolder/File Naming Convention: [duration]/100y.csv, e.g., 24h/100y.csv
Data Dimension: 207,173 (R) x 3 (C)
Unit: %
Format: C1: Latitude; C2: Longitude; C3: Bias in the 100-year event based on PREC-IDF vs. NG-IDF

SI/

Description: Seasonality of annual maximum W events with 24-, 48-, and 72-h durations over water years 1951‒2013
Subfolder/File Naming Convention: [duration]/W.csv, e.g., 24h/W.csv
Data Dimension: 207,173 (R) x 4 (C)
Format: C1: Latitude; C2: Longitude; C3: Mean data (=1 if Oct 1); C4: Seasonality index ranging from 0 to 1