search

Freshwater-Initiative / SkagitLandslideHazards

Seattle City Light is interested in improving understanding of landslide hazard and sediment transport to ensure reliable and cost-effective hydropower generation.
4 stars 4 forks source link

DHSVM input climatology #2

Open ChristinaB opened 6 years ago

ChristinaB commented 6 years ago
RondaStrauch commented 6 years ago

Can we produce a map of peak snowmelt?

ChristinaB commented 5 years ago

These maps are all in the Skagit report. The digitial appendix for the skagit report has all this info.

Do we want to add anything to the Appendix ? Or just reference the Skagit report?

ChristinaB commented 5 years ago

Email to SeYeun Lee Jun 28, 2018

All of the historic and future climate forcings are uploaded on HydroShare on this resource link: https://www.hydroshare.org/resource/6b81ea41c648413981c7f6a82a728a5a/

This also has some draft climate figures for the report. I'm not sure which we will want to use, or which additional ones we want to be made for future climate.

I also put them in Pogolinux./data1/mount_sea/SaukDHSVM_2018/forcings and uncompressed everything so it is ready to go.

If there are access issues or you want to run them elsewhere, run this script to wget each tar file

This is all 20 models (10 GCM and RCP 45 and 85). If you want only three, we can look at the model streamflow results from the full Skagit extent: https://www.hydroshare.org/resource/f4a060f538184d49b642b7c079cbe0be/ For more locations, I gave you view permissions to the private HydroShare resource of previous model outputs that are under review by SCL and USIT. These can not be distributed until after the USIT has finished their review.

I know you usually choose GCM comparisons by P and T but I do not have that analysis. Below are three models you could compare to explore the range of high and low flows predicted. These are from the GCM tab in the attached Sauk near Sauk model output.

From low flow statistics RCP85 2060-2099

Wettest Summer Model | NorESM1-M Mid | CNRM-CM5 Driest Summer Model | HadGEM2-ES365   |   From Peak Annual Events 2010-2099 Highest Annual Flood | CCSM4 Mid | IPSL-CM5A-MR Low Annual Flood | CNRM-CM5   |   From 5% Highest Daily Flow November  Statistics RCP85 2060-2099 High Flows Nov | CSIRO-Mk3-6-0 Mid | CCSM4 Low Flows Nov | MIROC5   |   From 5% Highest Daily Flow June Statistics RCP85 2060-2099 High Flows June | NorESM1-M Mid | bcc-csm1-1-m Low Flows June | HadGEM2-ES365

ChristinaB commented 5 years ago

SaukRiver_nrSauk_draft_future_GCManalysis.xlsx This is the spreadsheet with the annual peak flows for the Sauk River used to select GCM's for the Sauk based on October Exc Prob = 0.9 for the end of century average

  ExcProb/cfs/rcp85/2060-2099 Oct
  GCM P0.9
1 HadGEM2-ES365 135.4979
2 CSIRO-Mk3-6-0 150.3982
3 HadGEM2-CC365 160.9266
4 CanESM2 164.8261
5 CNRM-CM5 174.6592
6 IPSL-CM5A-MR 180.7942
7 CCSM4 189.798
8 MIROC5 195.4297
9 bcc-csm1-1-m 241.39
10 NorESM1-M 272.3251

This is how we selected the GCMs to use for stream temperature modeling in the Sauk to be sure we capture a general range.

ChristinaB commented 5 years ago

Because we care about high flow/saturation events we looked at how the sediment yield (controlled by distribution of highest flow events) ranges for these models. We want to focus on the 2050s with 2025-2075. The three models that capture the range of low flows based on Oct 0.9 Exc Prob also capture the range of sediment yield. IPSL was VERY high, and BCC and HadGEM-CC365 are also (suspiciously higher at the end of century.

image

RondaStrauch commented 5 years ago

SCL Appendix 12 from 2016 IRP selected these 3 GCMs for analyzing streamflow (supply) and demand (heating).

image
RondaStrauch commented 5 years ago

Noticed that we used HadGEM2-CC above, but in the bar graph above, this model produces very large sediment yields later in the century. Thus, we used HadGEM2-ES because this was relatively close to the same mid flow and most warming and had high sediment load at mid century.

ChristinaB commented 5 years ago

To be sure that the estimate from Sauk studies is relevant in the Skagit study area upstream of the Sauk, we looked at streamflow model results at these locations:

Comparing all GCMs at 2025-2074 the minimum for Percent Change of Highest Daily Flow, 100-year recurrence intervale, for 4.5 and 8.5, to see which GCMs generate the min and max % change for extreme high streamflow conditions.

model 2025-2074 Percent change in Q100 RCP 4.5
NorESM1-M RCP 4.5 -12.9026 % min GCM
CNRM-CM5 RCP 4.5 1.137662 mid GCM
HadGEM2-ES365 RCP 4.5 39.0274 % max GCM
model 2025-2074 Percent change in Q100 RCP 8.5
CNRM-CM5 RCP 8.5 -31.1485 % min GCM
NorESM1-M RCP 8.5 -18.96 % mid GCM
HadGEM2-ES365 RCP 8.5 19.46695 % max GCM

Of both scenarios and all models, HadGEM2-ES365 RCP 4.5 39.0274 % is the max GCM. and CNRM-CM5 RCP 8.5 -31.1485 % is the min GCM. NorESM is the wettest model in terms of annual average streamflow (based on Ross dam analysis).