Support the Arctic Sea Ice Forum and Blog

Author Topic: RASM-ESRL Experimental Sea Ice Forecasts  (Read 7670 times)

Michael

  • New ice
  • Posts: 49
    • View Profile
  • Liked: 0
  • Likes Given: 0
RASM-ESRL Experimental Sea Ice Forecasts
« on: October 06, 2017, 10:20:03 AM »
Quote
These 0-10 day, experimental, sea ice forecasts are produced by the NOAA Physical Sciences Division from a fully coupled ice-ocean-atmosphere model called RASM-ESRL. RASM-ESRL is run daily and posted online at 2 UTC. The model is initialized with the NOAA Global Forecast System (GFS) analyses and the Advanced Microwave Scanning Radiometer 2 (AMSR2) sea ice concentrations. The model is forced at the lateral boundaries by 3-hourly GFS forecasts of winds, temperature, and water vapor.
Citation

Quote
Plot provided by the NOAA/ESRL Physical Sciences Division, Boulder, CO from their Web site at http://www.esrl.noaa.gov/psd/.

url : https://www.esrl.noaa.gov/psd/forecasts/seaice/index.html

Michael

  • New ice
  • Posts: 49
    • View Profile
  • Liked: 0
  • Likes Given: 0
Re: RASM-ESRL Experimental Sea Ice Forecasts
« Reply #1 on: October 06, 2017, 10:23:03 AM »
The September minimum provides the idea opportunity to test the RASM_ESRL forecast in terms of sea ice extent, area and volume.

The REB.2017-09-dd.nc files provide a ten day forecast at 6 hourly steps. The first time step covers the time period 00:00 to 06:00, although it is not clear if this reflects the initial conditions or is a forecast for this period. It is this time period that is tracked from when it first appears at time step 36 until it finally appears as time step 0.

Below is tabulated the RASM_ESRL forecast for September Sea Ice Extent, for sea ice with a concentration greater than 15%.

The final column is the difference between the first forecast and the final value for that day.

Date_model time_Time step_Time step_Time step_Time step_Time step_Time step_Time step_Time step_Time step_Time step__Time step
3632282420161284036 - 0
2017-09-01736448.254.64.57554.56884.60134.61444.60964.61344.45334.45244.45310.1469
2017-09-02736449.254.57414.57634.61824.61494.60044.6064.44384.44224.44284.44220.1319
2017-09-03736450.254.57894.62624.6144.58674.59334.42634.42924.43664.43334.43640.1425
2017-09-04736451.254.6284.61694.56714.58474.41484.41794.43154.42464.42574.42590.2021
2017-09-05736452.254.6034.54634.57594.38774.41124.4224.41814.42554.4284.42790.1751
2017-09-06736453.254.53684.5724.37454.40794.41424.40944.42044.42064.42434.32250.2143
2017-09-07736454.254.58094.36634.40374.40774.40024.41994.41684.41834.33934.33910.2418
2017-09-08736455.254.36644.40674.38494.44.41934.40424.4144.33614.3414.34160.0248
2017-09-09736456.254.424.36414.39634.4214.40814.40924.33974.34194.34374.34350.0765
2017-09-10736457.254.35154.39714.42924.42244.424.35164.35784.3534.35344.354-0.0025
2017-09-11736458.254.40884.444.444.42984.36114.36354.36464.36854.36784.36730.0415
2017-09-12736459.254.45554.44024.43074.36444.37174.36514.37384.37354.3765
2017-09-13736460.254.44924.43594.36524.37124.35814.37244.37354.37564.2240.2252
2017-09-14736461.254.44734.36974.37164.36574.37784.38514.38174.23984.2390.2083
2017-09-15736462.254.38414.36274.37694.38174.39384.4024.26164.26084.26040.1237
2017-09-16736463.254.3524.38014.38354.40254.41254.28334.28524.2774.22630.1257
2017-09-17736464.254.37054.38734.40144.42584.29454.2974.29774.25464.25480.1157
2017-09-18736465.254.40344.40854.43584.30284.3094.32074.27344.28154.28110.1223
2017-09-19736466.254.41754.43854.33274.31894.3374.29074.30194.30694.38210.0354
2017-09-20736467.254.43984.34944.35234.37334.35064.38284.38734.44544.526-0.0862
2017-09-21736468.254.35394.39684.43374.42314.44254.44354.49614.55294.554
2017-09-22736469.254.36284.42074.47734.45984.46754.46794.5244.57664.57794.5783-0.2155
2017-09-23736470.254.4384.48974.46884.47474.48194.53914.58554.59224.59114.592-0.154
2017-09-24736471.254.5014.46694.47464.49784.55714.59574.60874.60554.61174.6121-0.1111
2017-09-25736472.254.47294.48154.51364.57944.62944.63354.63944.65134.65774.6558-0.1829
2017-09-26736473.254.49684.53854.62634.67344.68224.69634.70214.71044.71094.711-0.2142
2017-09-27736474.254.56714.67434.71194.72034.74084.75244.76444.76474.76144.7221-0.155
2017-09-28736475.254.70134.75324.75814.79274.82194.8314.8194.81844.83044.8312-0.1299
2017-09-29736476.254.7894.79624.82364.87614.8844.85894.86024.86874.87464.8747-0.0857
2017-09-30736477.254.84914.86164.91784.93644.88564.89764.90334.90764.9064.9042-0.0551

For convenience I am splitting this post into three. Sea Ice Area follows...

Michael

  • New ice
  • Posts: 49
    • View Profile
  • Liked: 0
  • Likes Given: 0
Re: RASM-ESRL Experimental Sea Ice Forecasts
« Reply #2 on: October 06, 2017, 10:24:16 AM »
...Below is tabulated the RASM_ESRL forecast for September Sea Ice Area, for sea ice with a concentration greater than 15%.

The final column is the difference between the first forecast and the final value for that day.

Date_model time_Time step_Time step_Time step_Time step_Time step_Time step_Time step_Time step_Time step_Time step__Time step
3632282420161284036 - 0
2017-09-01736448.253.77243.74913.75393.76223.77553.77423.77823.6533.65523.65560.1168
2017-09-02736449.253.73953.74583.75413.76743.76583.77053.64173.64473.64623.64640.0931
2017-09-03736450.253.74023.74723.76043.75923.76393.63173.63523.63753.63863.63780.1024
2017-09-04736451.253.74123.75443.7533.75593.62253.62343.62733.62793.62863.62910.1121
2017-09-05736452.253.74463.74423.74343.60633.6143.61943.6193.62123.62383.62390.1207
2017-09-06736453.253.73583.72993.58993.60593.61473.61353.6163.6193.62133.7573-0.0215
2017-09-07736454.253.72193.5793.59633.60993.60973.61073.61493.61863.73883.7394-0.0175
2017-09-08736455.253.56933.58673.60123.60733.60513.60953.61623.73093.73573.7362-0.1669
2017-09-09736456.253.57973.59293.60383.6023.60933.61513.7283.73433.73833.7386-0.1589
2017-09-10736457.253.58753.60133.59973.61023.6153.72433.73123.73673.73993.7404-0.1529
2017-09-11736458.253.6033.59723.61343.61473.72073.7273.73243.73533.73773.738-0.135
2017-09-12736459.253.59713.62263.61823.71723.72633.7263.72963.73373.7374
2017-09-13736460.253.6393.62483.71623.7263.72773.73053.73563.73883.7638-0.1248
2017-09-14736461.253.63593.71753.72653.73753.73253.73273.74623.75023.751-0.1151
2017-09-15736462.253.72023.72263.74673.73273.73623.75793.74593.75453.7546-0.0344
2017-09-16736463.253.71953.74723.73753.7433.773.75463.76563.76343.5960.1235
2017-09-17736464.253.73773.74943.7443.77063.77213.77853.77943.63713.63720.1005
2017-09-18736465.253.76183.74943.77493.78423.7953.80723.68933.69993.70020.0616
2017-09-19736466.253.75823.78913.80143.81323.82173.72733.74063.75013.7975-0.0393
2017-09-20736467.253.80083.82163.85523.83383.76673.78653.79893.86974.0958-0.295
2017-09-21736468.253.83123.90973.8693.82083.83993.85713.91764.12934.1302
2017-09-22736469.253.83053.93163.90963.85273.86443.8843.9414.13684.14324.144-0.3135
2017-09-23736470.253.94723.92693.8633.87083.89433.94774.13084.13974.14754.1487-0.2015
2017-09-24736471.253.93923.86673.87183.90513.95894.12484.13954.15454.16274.1637-0.2245
2017-09-25736472.253.87333.88093.92273.98264.13084.15084.1744.18294.20044.2007-0.3274
2017-09-26736473.253.90723.94754.01674.15434.16944.20814.21984.24384.25564.2554-0.3482
2017-09-27736474.253.99334.06464.18084.19474.24014.26394.29854.30674.31624.3572-0.3639
2017-09-28736475.254.11374.21614.23714.29634.33644.36744.37274.38244.41744.4172-0.3035
2017-09-29736476.254.26214.28754.34854.41634.43564.42474.4444.4644.48824.4884-0.2263
2017-09-30736477.254.33954.39224.46114.51624.46694.49294.50024.52554.52954.5294-0.1899

Sea Ice Volume follows...

Michael

  • New ice
  • Posts: 49
    • View Profile
  • Liked: 0
  • Likes Given: 0
Re: RASM-ESRL Experimental Sea Ice Forecasts
« Reply #3 on: October 06, 2017, 10:34:51 AM »
...Below is tabulated the RASM_ESRL forecast for September Sea Ice Volume.

The final column is the difference between the first forecast and the final value for that day.

Date_model time_Time step_Time step_Time step_Time step_Time step_Time step_Time step_Time step_Time step_Time step__Time step
3632282420161284036 - 0
20170901736448.257.29217.27197.28097.26537.25067.25177.25886.99286.96726.950.3421
20170902736449.257.26297.26457.2427.23217.23667.24796.98496.9646.95446.94580.3171
20170903736450.257.24877.22287.21977.23227.24146.97846.95976.95126.94936.9380.3107
20170904736451.257.20867.20367.23467.23416.97196.9496.94576.94616.94266.93810.2705
20170905736452.257.18557.23467.21716.96276.93666.94686.94546.94396.94826.92830.2572
20170906736453.257.23287.19456.94436.92676.94266.94546.94356.95436.93917.10920.1236
20170907736454.257.17096.93786.91636.93586.93796.93126.95066.94017.10997.09860.0723
20170908736455.256.93316.90046.92686.92796.91586.94036.93357.1067.10937.1096-0.1765
20170909736456.256.88216.91786.91846.89596.92666.92437.09947.10857.12297.1165-0.2344
20170910736457.256.91226.96.86566.90946.90427.07557.08697.10577.10747.0965-0.1843
20170911736458.256.90176.84626.90366.88637.05147.06497.08327.08377.08377.0703-0.1686
20170912736459.256.83066.91136.88177.02927.05287.06517.06767.07227.0689
20170913736460.256.92636.897.01777.05027.07077.07067.07327.07177.0347-0.1084
20170914736461.256.90667.00547.04887.07867.06057.05167.07817.03657.0241-0.1175
20170915736462.257.00117.04187.07877.05127.04447.07997.02337.03217.0285-0.0274
20170916736463.257.0327.06777.05717.03837.08417.02987.04747.04755.96511.0669
20170917736464.257.04977.06827.03667.0747.04587.06117.06685.99175.9911.0587
20170918736465.257.06677.04877.0797.05077.07697.09596.03656.05416.05641.0103
20170919736466.257.06067.10237.0757.10917.10046.06476.08686.10446.11750.9431
20170920736467.257.13167.10557.18957.09376.07126.08376.10526.13946.29050.8411
20170921736468.257.11667.2457.12126.06676.07126.09626.12436.25886.269
20170922736469.257.09837.25087.14796.05376.06046.08216.10466.226.246.24490.8534
20170923736470.257.25417.14976.05016.0546.06436.0816.19276.2186.23166.23641.0177
20170924736471.257.15746.05186.0526.05496.06366.16646.19556.22016.23166.23850.9189
20170925736472.256.05246.05646.05966.06676.1466.18326.21386.22366.24376.2511-0.1987
20170926736473.256.07616.06586.08366.14186.17426.20746.22046.24696.27296.2807-0.2046
20170927736474.256.09716.11496.13766.16756.19446.21496.24626.27456.29736.4698-0.3727
20170928736475.256.16096.14536.18196.2086.22836.25856.28966.30836.46976.4812-0.3203
20170929736476.256.15986.19286.23336.26246.28166.316.33526.46996.49856.5145-0.3547
20170930736477.256.20686.26046.27876.31486.32026.35396.46686.4976.51566.5297-0.3229


Michael

  • New ice
  • Posts: 49
    • View Profile
  • Liked: 0
  • Likes Given: 0
Re: RASM-ESRL Experimental Sea Ice Forecasts
« Reply #4 on: October 06, 2017, 10:35:56 AM »
The RASM_ESRL model uses a 10 km polar stereoscopic grid with a Standard Latitude of 84.5N, on a spherical Earth with a radius of 6370 km.

The RASM_ESRL grid is reported as an Arakawa B-grid (scalars at centre, vectors at NE corner), but what relationship this has with the POP/CICE Arakawa B-grid is unclear. Certainly they are not coincident, the RASM_ESRL grid varies in grid cell size from 9.364 to 10.023 km while the maximum size of the POP/CICE grid is 9.264 km. Also the 1/12 degree mesh used by POP/CICE appears to have an "equator" around the 157/337 degree meridian. As the RASM_ESRL model is a synthesis of several models, each with different properties, it would seem more sensible to use an Arakawa A-grid with scalars and vectors at the centre of the grid cells for the RASM_ESRL model.

The choice of a spherical earth grid is also curious. If the model is to have a real world use it would seem more logical to use a grid based on the WGS84 ellipsoid.

Michael

  • New ice
  • Posts: 49
    • View Profile
  • Liked: 0
  • Likes Given: 0
Re: RASM-ESRL Experimental Sea Ice Forecasts
« Reply #5 on: October 06, 2017, 10:40:13 AM »
Below is a summary of RASM_ESRL Sea Ice for September 2017.

As there is no official "nowcast" the first time step in the REB.2017-09-dd.nc files is assumed to be representative of that day.

Extent and Area are calculated for sea ice with a concentration greater than 15%.

The Effective Thickness in the final column is the average effective thickness for sea ice thicker than 0.15 m. This figure is directly comparable with the PIOMAS Average Thickness chart.

There is no data file for 12 September 2017.

Units : Extent and Area - 106 km2, Volume - 103 km3, Thickness - m

Date_model time _Extent_Area_Volume_Average__*Effective
thicknessthickness*
2017-09-01736448.254.45313.65566.951.88961.5048
2017-09-02736449.254.44223.64646.94581.89331.5047
2017-09-03736450.254.43643.63786.9381.89571.5032
2017-09-04736451.254.42593.62916.93811.89991.5038
2017-09-05736452.254.42793.62396.92831.89991.4998
2017-09-06736453.254.32253.75737.10921.88351.6298
2017-09-07736454.254.33913.73947.09861.88951.6143
2017-09-08736455.254.34163.73627.10961.89421.614
2017-09-09736456.254.34353.73867.11651.89461.6123
2017-09-10736457.254.3543.74047.09651.88841.6016
2017-09-11736458.254.36733.7387.07031.88271.5888
2017-09-12736459.25
2017-09-13736460.254.2243.76387.03471.8611.6587
2017-09-14736461.254.2393.7517.02411.86391.6394
2017-09-15736462.254.26043.75467.02851.86381.6336
2017-09-16736463.254.22633.5965.96511.6511.4457
2017-09-17736464.254.25483.63725.9911.63971.4464
2017-09-18736465.254.28113.70026.05641.62951.4581
2017-09-19736466.254.38213.79756.11751.60471.4571
2017-09-20736467.254.5264.09586.29051.52981.5211
2017-09-21736468.254.5544.13026.2691.51211.5033
2017-09-22736469.254.57834.1446.24491.50141.4869
2017-09-23736470.254.5924.14876.23641.49761.4781
2017-09-24736471.254.61214.16376.23851.49241.4701
2017-09-25736472.254.65584.20076.25111.48261.463
2017-09-26736473.254.7114.25546.28071.4711.4606
2017-09-27736474.254.72214.35726.46981.47991.4956
2017-09-28736475.254.83124.41726.48121.46181.4857
2017-09-29736476.254.87474.48846.51451.44591.4828
2017-09-30736477.254.90424.52946.52971.43641.4779

A-Team

  • Young ice
  • Posts: 2977
    • View Profile
  • Liked: 944
  • Likes Given: 35
Re: RASM-ESRL Experimental Sea Ice Forecasts
« Reply #6 on: October 06, 2017, 05:07:45 PM »
Quote
no official "nowcast" the first time step in the REB.2017-09-dd.nc files is assumed  ...seem more sensible to use an Arakawa A-grid with scalars and vectors at the centre of the grid cells for the RASM_ESRL model.... seem more logical to use a grid based on the WGS84 ellipsoid.... hat relationship this has with the POP/CICE Arakawa B-grid
So what did they say when you wrote them with these questions/suggestions?

They solicit feedback, emails are shown. There are other issues, like going on with the 'Alaska-centric' view on some of their daily offerings, rather than whole Arctic Ocean plus Barents view that they announced back in mid-August. Some of the static graphic pairs are mis-sized. Contouring is over-writing binned visualizations, unnecessary. No data files provided for some graphics, not easy to duplicate.

Panoply updated this morning to 4.8.4 (windows, mac, linux) for those wishing to make maps from REB.2017-09-05 or RASM-ESRL_2017-10-05.

https://www.giss.nasa.gov/tools/panoply/

Michael

  • New ice
  • Posts: 49
    • View Profile
  • Liked: 0
  • Likes Given: 0
Re: RASM-ESRL Experimental Sea Ice Forecasts
« Reply #7 on: October 07, 2017, 10:40:59 AM »
So what did they say when you wrote them with these questions/suggestions?

They were very helpful, but reticent.

Helpful in providing the details of their grid so that I could generate the figures for extent etc.

They are very reticent when it comes to the CICE model. The team running RASM is based at the Naval Postgraduate School, and their website is an official Navy website. So there are possible security issues involved. The RASM team are currently working on implementing a 1/48  degree mesh for CICE, which will give it a finer resolution than the 3.125 km that HYCOM-CICE achieves with a tripole grid.

The RASM-ESRL team appear to be open to suggestions, but as the project seems to be under development there will probably be changes as things settle down.

Michael

  • New ice
  • Posts: 49
    • View Profile
  • Liked: 0
  • Likes Given: 0
Re: RASM-ESRL Experimental Sea Ice Forecasts
« Reply #8 on: October 08, 2017, 07:44:23 PM »
Sea ice concentration is fundamental to the operation of the CICE element of the RASM-RSRL model, and as it is initialized with ASMR2 a comparison with Uni Hamburg sea ice concentration seems justified.

RASM-ESRL data for the period 00:00 to 06:00 7 Oct 2017 is used for the comparison. The Uni Hamburg 3.125 km data have been regridded onto the RASM-ESRL  10 km grid.

I have used the HYCOM CICE colour scales and have included daily RASM-ESRL sea ice concentration and thickness.

* Click to animate GIF*
« Last Edit: November 02, 2017, 11:35:58 AM by Michael »

Michael

  • New ice
  • Posts: 49
    • View Profile
  • Liked: 0
  • Likes Given: 0
Re: RASM-ESRL Experimental Sea Ice Forecasts
« Reply #9 on: October 27, 2017, 09:15:08 AM »
@ Neven :  I had the same problem previously reported with .PNG images with the above animated .GIF.

Neven

  • Administrator
  • First-year ice
  • Posts: 9470
    • View Profile
    • Arctic Sea Ice Blog
  • Liked: 1333
  • Likes Given: 617
Re: RASM-ESRL Experimental Sea Ice Forecasts
« Reply #10 on: October 27, 2017, 09:40:32 AM »
We're looking into it, Michael. I'll report there that you experience the same problem with an animated GIF.
The enemy is within
Don't confuse me with him

E. Smith

DungeonMaster

  • Administrator
  • Frazil ice
  • Posts: 152
    • View Profile
  • Liked: 18
  • Likes Given: 3
Re: RASM-ESRL Experimental Sea Ice Forecasts
« Reply #11 on: November 01, 2017, 07:48:34 PM »
This problem with images should be fixed now (probably until next version, I'll reapply it when necessary). Sorry for the delay.


And thanks for your posts!
This forum helps me to feel less uncomfortable about "doing something" about the melting Arctic and the warming world.

Michael

  • New ice
  • Posts: 49
    • View Profile
  • Liked: 0
  • Likes Given: 0
Re: RASM-ESRL Experimental Sea Ice Forecasts
« Reply #12 on: November 04, 2017, 10:47:08 AM »
Rather belatedly, here is a summary of RASM_ESRL Sea Ice for October 2017.

I have used the REB.2017-10-dd.nc files, and taken time step 0 as being the "nowcast" and time step 39 as being the day 10 forecast.

Extent and Area are calculated for sea ice with a concentration greater than 15%.

The Effective Thickness in the final column is the average effective thickness for sea ice thicker than 0.15 m. This figure is directly comparable with the PIOMAS Average Thickness chart.

Units : Extent and Area - 10^6 km^2, Volume - 10^3 km3, Thickness - m
Date_model time _Extent_Area_Volume_Average__Maximum__*Effective
thicknessthicknessthickness*
20171001736478.254.94134.56596.54351.4283.60761.4714
20171002736479.254.97374.60286.56181.42063.59981.4666
20171003736480.255.00484.6356.57031.41263.59441.4591
20171004736481.255.10154.76816.71451.40413.61021.4567
20171005736482.255.18154.80866.72431.39343.60731.4506
20171006736483.255.27534.88326.7441.37533.6051.4442
20171007736484.255.39124.96756.75791.35533.60331.431
20171008736485.255.46515.04766.77671.33823.60171.4193
20171009736486.255.51825.11696.79881.32453.59811.4077
20171010736487.255.57515.18026.83221.31483.59321.3962
20171011736488.255.64375.23826.57411.25053.58951.3446
20171012736489.255.82435.38776.60871.22223.5841.3362
20171013736490.255.93325.49716.64381.20483.57561.3263
20171014736491.256.00145.57976.67841.19333.5711.3077
20171015736492.256.07325.66256.71341.1823.56851.289
20171016736493.256.16645.75156.74291.16893.56611.2683
20171017736494.256.29695.8586.7751.15343.56351.2484
20171018736495.256.27285.88166.85261.16023.56261.2564
20171019736496.256.68616.19256.91441.11423.56051.2369
20171020736497.256.79476.4286.991.08523.55941.2229
20171021736498.256.88676.56257.05041.07233.55811.2114
20171022736499.256.98226.57647.08261.07493.55451.1706
20171023736500.257.06496.65317.13741.07083.55221.1454
20171024736501.257.13296.72857.19881.06793.55071.1286
20171025736502.257.05766.66497.27261.08853.55031.1394
20171026736503.257.32636.89837.36671.06623.54651.1312
20171027736504.257.40487.04947.45281.05563.54411.1197
20171028736505.257.47167.1167.52451.05573.54381.1094
20171029736506.257.53057.16387.59251.0583.54351.1007
20171030736507.257.56587.19357.63611.05973.53831.0964
20171031736508.257.60677.23327.70331.06323.53811.0957

It is notable that thickness decreases during the month. This is possibly because the RASM-ESRL model is not manufacturing enough sea ice to keep pace with the increase in extent.

Below are animations showing the progress of the sea ice thickness and concentration "nowcasts", and a comparison of RASM-ESRL sea ice concentration with the University of Hamburg ASI AMSR2 sea ice concentration.

Click to animate

echoughton

  • Frazil ice
  • Posts: 121
    • View Profile
  • Liked: 21
  • Likes Given: 82
Re: RASM-ESRL Experimental Sea Ice Forecasts
« Reply #13 on: November 04, 2017, 11:51:11 AM »
Isn't the decrease in thickness for the month a consequence of all the thin extent growth for early in the season? If we have new widespread extent, it certainly will be thin to start, lowering the the overall thickness average,

Michael

  • New ice
  • Posts: 49
    • View Profile
  • Liked: 0
  • Likes Given: 0
Re: RASM-ESRL Experimental Sea Ice Forecasts
« Reply #14 on: November 05, 2017, 11:29:53 AM »
Isn't the decrease in thickness for the month a consequence of all the thin extent growth for early in the season? If we have new widespread extent, it certainly will be thin to start, lowering the the overall thickness average,
You are probably correct, I was misled by the reduction in the maximum reported thickness.

The CPOM CryoSat-2 sea ice thickness data have been available since  8 October, so here is a comparison between RASM-ESRL and the CPOM 2-day, 1-km product. The difference is calculated as RASM-ESRL minus CPOM thickness.