Kimberly Smith

University of Utah

 Recent Activity

ABSTRACT:

This collection of MATLAB scripts runs the temperature component of the stochastic harmonic autoregressive parametric (SHArP) weather generator, which simulates trended, nonstationary precipitation occurrence, precipitation amount, and temperature values. The precipitation component is not included here because it is based off of previous weather generators and doesn't use any new methods. The temperature component follows a concept introduced in previous studies, but SHArP simulates temperature values directly as opposed to simulating the temperature residuals. SHArP weather generator validation and illustrations at a single site have been published in the Journal of Applied Meteorology and Climatology (Smith et al. 2017; "A new method for generating stochastic simulations of daily air temperature for use in weather generators"; https://doi.org/10.1175/JAMC-D-16-0122.1), and the multisite generalization of SHArP is in press in the same journal (Smith et al. 2018; "Multisite generalization of the SHArP weather generator"; https://doi.org/10.1175/JAMC-D-17-0236.1).

How to use:
- All input data need to be combined into a single .mat structure file and contain the number of locations to be simulated, the number of days to simulate, a time matrix (in format: year month day), the day of year corresponding to each day, the precipitation occurrence 1s and 0s in logical format corresponding to "wet" or "dry" days (determined from observations or downscaled climate model output), the maximum temperature training data for each site (i.e. observations or downscaled climate model output), and the minimum temperature training data for each site.
- This .mat structure file (called "sett") is the input to fit_SHArP_1, which fits the parameters needed to run the model. fit_SHArP_1 internally uses fit_SHArP_var_1a to determine the coefficients for each of the 26 variables per site in the model; this latter script is not run separately. The output from fit_SHArP_1 is called the driver_t structure and needs to contain maximum and minimum temperature data (from observations or downscaled climate model output), the precipitation occurrence 1s and 0s in logical format corresponding to "wet" or "dry" days (determined from observations or downscaled climate model output), a time matrix (in format: year month day), a variable structure that contains the 26 coefficients for each variable at each site, b_k, c_k, and the results from the EOF analysis on the precipitation occurrence patterns.
- The driver_t structure is the input to simulate_SHArP_2, which gives maximum temperature and minimum temperature as output. This script can also be modified to output the temperature mean and noise but note that this will increase computation time significantly.

Show More

 Contact

Resources
All 0
Collection 0
Composite Resource 0
Generic 0
Geographic Feature 0
Geographic Raster 0
HIS Referenced Time Series 0
Model Instance 0
Model Program 0
MODFLOW Model Instance Resource 0
Multidimensional (NetCDF) 0
Script Resource 0
SWAT Model Instance 0
Time Series 0
Web App 0
Script Resource Script Resource
Stochastic Harmonic Autoregressive Parametric (SHArP) Weather Generator
Created: June 25, 2018, 8:21 p.m.
Authors: Kimberly Smith · Courtenay Strong · Firas Rassoul-Agha

ABSTRACT:

This collection of MATLAB scripts runs the temperature component of the stochastic harmonic autoregressive parametric (SHArP) weather generator, which simulates trended, nonstationary precipitation occurrence, precipitation amount, and temperature values. The precipitation component is not included here because it is based off of previous weather generators and doesn't use any new methods. The temperature component follows a concept introduced in previous studies, but SHArP simulates temperature values directly as opposed to simulating the temperature residuals. SHArP weather generator validation and illustrations at a single site have been published in the Journal of Applied Meteorology and Climatology (Smith et al. 2017; "A new method for generating stochastic simulations of daily air temperature for use in weather generators"; https://doi.org/10.1175/JAMC-D-16-0122.1), and the multisite generalization of SHArP is in press in the same journal (Smith et al. 2018; "Multisite generalization of the SHArP weather generator"; https://doi.org/10.1175/JAMC-D-17-0236.1).

How to use:
- All input data need to be combined into a single .mat structure file and contain the number of locations to be simulated, the number of days to simulate, a time matrix (in format: year month day), the day of year corresponding to each day, the precipitation occurrence 1s and 0s in logical format corresponding to "wet" or "dry" days (determined from observations or downscaled climate model output), the maximum temperature training data for each site (i.e. observations or downscaled climate model output), and the minimum temperature training data for each site.
- This .mat structure file (called "sett") is the input to fit_SHArP_1, which fits the parameters needed to run the model. fit_SHArP_1 internally uses fit_SHArP_var_1a to determine the coefficients for each of the 26 variables per site in the model; this latter script is not run separately. The output from fit_SHArP_1 is called the driver_t structure and needs to contain maximum and minimum temperature data (from observations or downscaled climate model output), the precipitation occurrence 1s and 0s in logical format corresponding to "wet" or "dry" days (determined from observations or downscaled climate model output), a time matrix (in format: year month day), a variable structure that contains the 26 coefficients for each variable at each site, b_k, c_k, and the results from the EOF analysis on the precipitation occurrence patterns.
- The driver_t structure is the input to simulate_SHArP_2, which gives maximum temperature and minimum temperature as output. This script can also be modified to output the temperature mean and noise but note that this will increase computation time significantly.

Show More