Surface melt predictions from Weather Research and Forecasting (WRF) model output for Greenland melt season, 1996-2005 and 2071-2080
Summary This dataset contains daily, Jun-Jul-Aug surface melt occurrence predictions made using near-surface temperatures from high resolution Polar Weather Research and Forecast (WRF) (Hines and Bromwich 2008; Skamarock et al 2008) simulations driven by European Centre for Medium-Range Weather Fore...
| Main Author: | |
|---|---|
| Format: | Dataset |
| Language: | English |
| Published: |
NSF Arctic Data Center
2020
|
| Subjects: | |
| Online Access: | https://dx.doi.org/10.18739/a2bz6189j https://arcticdata.io/catalog/view/doi:10.18739/A2BZ6189J |
| Summary: | Summary This dataset contains daily, Jun-Jul-Aug surface melt occurrence predictions made using near-surface temperatures from high resolution Polar Weather Research and Forecast (WRF) (Hines and Bromwich 2008; Skamarock et al 2008) simulations driven by European Centre for Medium-Range Weather Forecasts Re-Analysis (ERA) Interim (Dee et al, 2011; 1996-2005), the Community Earth System Model (CESM) Large Ensemble (Kay et al, 2015; 1996-2005 and 2071-2080) and the CESM Low Warming Ensemble (Sanderson et al 2016; 2071-2080). Along with the model output files, associated datasets include input files and scripts for preprocessing, model execution and making sample plots. Every effort has been made to provide all required input datasets for the model execution and plotting scripts but it is possible that missing files or other prereqs may exist. Changes will definitely be needed to customize scripts to the user’s runtime environment. Details The README.pdf in this dataset provides more details on the model used to predict surface melt from near-surface temperature. Briefly, satellite-based melting observations are used to partition temperatures into melt and “no melt” subsets. Distributions for these two temperature subsets show that temperatures fall into three categories: “no melt” (lowest), melt (highest) and an overlapping range where both melt and “no melt” were observed for the same temperature. The latter was converted to (1) a melt probability estimate (0-1) through a linear interpolation between distribution peaks and (2) a melt/”melt” estimate (0 or 1) by setting a temperature threshold that minimizes model error versus observations. 1. Melting prediction model input (a) WRF temperature data Temperature datasets are post-processed model output from the regional forecast model Polar WRF (i.e., WRF-Advanced Research WRF (ARW) with polar modifications developed by the Polar Meteorology Group at The Ohio State University; Hines and Bromwich 2008; Skamarock et al 2008). WRF data are on a 15 kilometers (km) spatial grid with 39 vertical levels and were originally archived every 3 hours for the core melt-season months of June, July and August. These 3-hourly data were averaged to daily to match surface melt observations. WRF initial and boundary conditions were provided by three global datasets: the ERA Interim reanalysis (European Centre for Medium-Range Weather Forecasts (ECMWF), erai, Dee et al, 2011), the CESM Large Ensemble (National Center for Atmospheric Research (NCAR), cesmle, Kay et al, 2015) and the CESM Low Warming Ensemble (National Center for Atmospheric Research (NCAR), cesmlw, Sanderson et al 2016). Simulations cover two broad time periods: historical (or hindcast) and future (based on standard emissions scenarios). Future, GCM-based simulations cover a “high warming” scenario based on RCP 8.5 and a “low warming” scenario based on limiting future temperature increases to 1.5 degrees Celsius above pre-industrial levels. These simulations are limited to time periods where the driving variables required by WRF were archived for the GCM at sub-daily resolution. CESM Large Ensemble-based datasets include ten ensemble members and an ensemble average. The CESM Low Warming ensemble archived sub-daily data needed by WRF for only one ensemble member (11). The ensemble average is thus effectively member 11. Model domain corner coordinates are: 55.2 North, 62.2 West (Southwest), 76.2 North, 117.2 West (Northwest), 76.2 North, 37.2 East (Northeast), 55.2 North, 17.8 West (Southeast). (b) Surface melt observations These data are from the “MEaSUREs Greenland Surface Melt Daily 25km Equal-Area Scalable Earth (EASE)-Grids 2.0, Version 1” (https://nsidc.org/data/NSIDC-0533/versions/1). Data were regridded from the 25 km EASE grid to match the WRF 15-km grid. The melt observations were also used to create an icesheet mask to constrain the melt prediction model to the current ice sheet extent. 2. Melt prediction model output For each time period, estimates of surface melt probability and surface melt occurrence were created using the prediction model. README.pdf has additional details on the model. 3. Scripts These datasets contain scripts for preprocessing, model execution and making sample plots. Many of these are Jupyter notebooks and thus require that Python tool. References Dee, D.P., with 35 co-authors, 2011: The ERA-Interim reanalysis: configuration and performance of the data assimilation system. Quart. J. R. Meteorol. Soc., 137, 553-597, doi: 10.1002/qj.828. Hines, K. M., and D. H. Bromwich, 2008: Development and testing of Polar WRF. Part I. Greenland ice sheet meteorology. Mon. Wea. Rev., 136, 1971-1989. Kay, J. E., with 20 co-authors, 2015: CESM Large Ensemble Project: A Community Resource for Studying Climate Change in the Presence of Internal Climate Variability, Bulletin of the American Meteorological Society, 96, 1333-1349, doi: 10.1175/BAMS-D-13-00255.1. Sanderson, B., B. O'Neill, and C. Tebaldi, 2016: What would it take to achieve the Paris temperature targets? Geophys. Res. Lett., doi: 10.1002/2016GL069563. Skamarock, W. C., Klemp, J. B., Dudhia, J., Gill, D. O., Barker, D., Duda, M. G., Huang, X.-Y., Wang, W., and Powers, J. G., 2008, A Description of the Advanced Research WRF Version 3 (No. NCAR/TN-475+STR). University Corporation for Atmospheric Research. doi:10.5065/D68S4MVH |
|---|