MegaMike

Edit - Misread that, but still informative The verification page is here: https://www.emc.ncep.noaa.gov/users/verification/ I recently found it while looking for model specifications. It looks like it's still "under construction." They have evaluations (surface and aloft) for the GFS, GEFS, SREF, and other climate modeling systems.

You're not looking at two instantaneous fields. edit: instantaneous not instantons lol Precipitation type and intensity: 6 hour, average precipitation rate... Honestly, I don't know why they'd choose to plot this. MSLP contours: instantaneous field interpolated by a programming utility. Use TT's 'MSLP & Precip (Rain/Frozen)' and 'Radar (Rain/Frozen)' graphics instead.

There are differences between 78' and 05', but I don't have time to describe them at the moment (I'm working, but I'm so distracted!). I ran semi-high resolution (d1 = 12km. d2 = 4km) WRF simulations for both events and both evaluated pretty/reasonably well:

The ICON has a pretty fine (approximated) resolution for a global model. From https://www.dwd.de/EN/research/weatherforecasting/num_modelling/01_num_weather_prediction_modells/icon_description.html: "the global ICON grid has 2,949,120 triangles, corresponding to an average area of 173 km² and thus to an effective mesh size of about 13 km." Small scale spatial/temporal errors magnify quite significantly past 84 hours... Look at the NAM as an example. Would you trust it past hour 84?

This is a good question. I conducted some evaluations on reanalysis and analysis data (consider it the initialization hour) for multiple modeling systems including the ERA5 (reanalysis system of the ECMWF), FNL (reanalysis system of the GFS), GFS (0.5x0.5deg), HRRR (3km), NAM (12km), and RAP (13km). All modeling systems predominately underestimate wind speed and overestimate wind gust. Keep in mind, I ran the evaluations for the entire month at 00, 06, 12, 18 UTC, therefore, the results are skewed in favor of fair weather conditions. I actually posted the results online for a couple reasons - When I ran evaluations operationally (WRF w/NAM and GFS forecast data), results were similar... Obviously, expect more erratic error past forecast hour 6.

Glad to share my thoughts and I absolutely agree! I still can't believe Pivotal requires payment for a product that hasn't been tested (Kuchera).

Ah! I didn't know that field was available. You are correct. In the GrADS Data Server, there is a field called, 'asnowsfc.' It looks like the field is obtained via HRRR's LSM (I'll double check later to be sure). You can view it here if you'd like: https://github.com/wrf-model/WRF/blob/master/phys/module_sf_ruclsm.F The AMS submission is here: https://journals.ametsoc.org/view/journals/wefo/25/1/2009waf2222311_1.xml?rskey=LFzwBc&result=4 To summarize its procedure, the weighted snow density (stepwise multiple linear regression) is first calculated by determining the (frozen) densities of snow, graupel, and sleet (as a function of the lowest atmospheric layer's temperature). The weighted snow density is then used to calculate SLR = (density of water) / (density of frozen). Snowfall is then the product of SWE * SLR. This is all done diagnostically using calls to the mp routine. It's also used to calculate new snowfall accumulation for its snow depth field. It's a pretty good algorithm actually compared to climatology and Kuchera. Taking the weighted contribution of different hydrometeors is a great method to avoid inflation of snowfall due to mixing. It's a huge advantage to have it calculated between modeled time steps too. Other modeling systems can incorporate this variable if the RUC LSM is applied.

11.5'' of very light snowfall in Wrentham, MA (measured at 11:30 AM). I'm teleworking in MA for now.

Let's try to clear up some things I've read: 1) Almost all NWP models (if not, all) provide liquid water equivalent (LWE) and snow water equivalent (SWE) as diagnostic fields. When I write, "diagnostic fields" this implies that both variables, LWE and SWE, is calculated within the model itself between modeled time steps. Modeled time steps are ~27 seconds depending upon the horizontal resolution of a modeling system. Regardless, the calculation of LWE is straightforward. LWE = the liquid accumulation of all hydrometeor contributions diagnosed by a modeling system The calculation of SWE varies by microphysics (mp) scheme. I'll give one example - one (there are many) mp scheme calculates fall rates for snow, pristine ice (I consider this sleet), graupel, and rain. Thus, fall rates can be converted into hydrometeor accumulation by multiplying by the modeled time step. Once the time stepped accumulation is obtained, SWE becomes - SWE = snow + sleet... Just keep in mind that SWE is often used to determine snowfall for all private/public websites. Since almost all mp schemes includes sleet within their routines, SWE may be overdone due to sleet contamination (for regions that mix). Thankfully, most websites make this clear. 2) mp schemes are reliable at fine grid scales (<3km). They can consider heat, moisture, and momentum flux caused by convection without parametrization. At larger scales (>12km), convective parameterizations are required to essentially approximate convictive processes. Although I'm not entirely sure, I believe cumulus parameterization from global models, to regional models, then to mesoscale models caused some forecasting problems for today's event. Just a thought! 3) The land surface model (LSM) of all NWP models includes a snow depth field that does determine new snowfall. Unfortunately, the algorithm is usually two-dimensional and the output isn't archived in gridded output... Delta snow depth (often provided by websites) will almost always be underdone since it includes gauge losses between the ground-snow and air-snow interface. 4) Anyone ever notice ''6-hour averaged Precip Rate" which is plotted by precipitation type? Think about it, does that make sense? No... They take an accumulated product (over the past 6 hours in this case), and use an instantaneous field to plot precipitation intensity by precipitation type. This is why you occasionally see snowfall along a cold front. Precipitation fell before the front passed (likely in a warm sector), BUT they use atmospheric fields post-frontal passage to determine precipitation type. This is no bueno imo. 5) Last one. Since snowfall is post-processed by most websites (excluding the ICON - I'd like to know what 'true SLR' means), keep in mind that they need instantaneous fields to calculate snow ratios. If output is provided every 6 hours, imagine how inaccurate snowfall may be when using instantaneous fields. I said this once, I'll say it again: Websites need to be descriptive. Especially as it relates to snowfall products.

It'll look very similar to the 12km NAM which it uses for its ic/bc's. I wouldn't trust it for any precipitation field anyways. The domain is too small. If winds are strong and perpendicular to any of its (4) boundaries, it'll muck up QPF along said boundary... I ran into this problem when I first ran WRF (ARW core) simulations. You can subtly see this flaw on the far southern boundary of their domain (within 5 grid points of their boundary). The stronger the winds, the worse it gets. I added a picture of a couple good examples of this artifact when winds are stronger, and more perpendicular, to a modeling system's domain.

I don't remember exactly the process, but several researches wrote a paper about the thermodynamic/microphysical evolution of that mega band: https://journals.ametsoc.org/view/journals/mwre/143/10/mwr-d-14-00407.1.xml?rskey=cIrzIt&result=2

You mean for the Boxing Day blizzard? I attached an interpolation that I created using GHCND/PNS observations. I created 200 other interpolations/images. I'm sure I'm not the only one who did this... Right? Anywho, I created the plots to determine the most significant snowfall event since 1900 on a grid. If you're looking for snowfall maps since 2008, I'd recommend checking out the NSA product: https://www.nohrsc.noaa.gov/snowfall/ They create their plots using in-situ observations, SLR climatology, Stage IV precipitation analysis (from Radar), and other gridded analysis products: https://www.weather.gov/media/notification/tins/tin15-05bigrsc_snowfall_aaa.pdf

I added more events. When I'm running/waiting on simulations at work, I run atmospheric simulations that'd interest me. I also added radar mosaics superimposed with observations and Atlantic basin, tropical disturbance storm tracks (by year) from ~1850 to 2020. This is done solely out of personal interest. I definitely used audio from several members of this forum, if you'd like me to remove the audio from the videos, let me know :). Otherwise, if you'd like me to run an additional simulation, give me a date and I'll consider initializing one. These are strictly deterministic (WRFv4.2) simulations with additional post-processing (unified post-processor/a Python library) utilities. Despite using analysis data (not forecasted data for the IC/BC's!!!), some events performed surprisingly poor ('Nemo' and the 'Boston Debacle' snowstorm to name a few). All events were evaluated at the surface and aloft when applicable. If you'd like to view the script I use to plot or run the simulations, let me know. I automated most tasks (.py>>.ncl>>.csh).

Hey guys, I've been running WRFv4.2.2 simulations for funzies lately. I configured WRFv4.2.2 w/1 parent (12km) and one nested (4km) domain. My nested domain (4km) encompasses all of New York City. If you'd like to view graphics of some notorious weather events, click the links below ... Keep in mind, my focal point is New England so some of these events weren't that interesting for the New York City area. Winter Storms (testing extratropical namelist options) Jan, 2005 - https://www.youtube.com/watch?v=Pe_75ngrc8Y&li st=PL7uw9vTkqQwp1s4oh6TRd8qm6StglCkUd&index=5 Apr, 1997 - https://www.youtube.com/watch?v=uEmCRmaUPtY&list=PL7uw9vTkqQwp1s4oh6TRd8qm6StglCkUd&index=2 Jan, 1996 - https://www.youtube.com/watch?v=J4o1Nz9wdBw&t=136s Mar, 1993 - https://www.youtube.com/watch?v=kyZWFzDvS4k&list=PL7uw9vTkqQwp1s4oh6TRd8qm6StglCkUd&index=3 Feb, 1978 - https://www.youtube.com/watch?v=sm3ipMFoHoE&list=PL7uw9vTkqQwp1s4oh6TRd8qm6StglCkUd Tropical Events (testing tropical namelist options) Irene - https://www.youtube.com/watch?v=GNhYYvrQrPc&t=252s Bob - https://www.youtube.com/watch?v=FmbhsaFaC4k&list=PL7uw9vTkqQwrX5nzMSy4qFQbTqU9WCqha&index=1 Gloria - https://www.youtube.com/watch?v=60wBgKfBrHg&list=PL7uw9vTkqQwrX5nzMSy4qFQbTqU9WCqha&index=3 Freezing Rain Events (testing p-type algorithm): Dec, 2008 - https://www.youtube.com/watch?v=sNkfVaudVZ8&list=PL7uw9vTkqQwrfqeW1diEdDmR_wGGvgjZC

I ran additional simulations if anyone's curious. All of them are at the following paths. Winter Storms (testing extratropical namelist options) Jan, 2005 - https://www.youtube.com/watch?v=Pe_75ngrc8Y&li st=PL7uw9vTkqQwp1s4oh6TRd8qm6StglCkUd&index=5 Apr, 1997 - https://www.youtube.com/watch?v=uEmCRmaUPtY&list=PL7uw9vTkqQwp1s4oh6TRd8qm6StglCkUd&index=2 Jan, 1996 - https://www.youtube.com/watch?v=J4o1Nz9wdBw&t=136s Mar, 1993 - https://www.youtube.com/watch?v=kyZWFzDvS4k&list=PL7uw9vTkqQwp1s4oh6TRd8qm6StglCkUd&index=3 Feb, 1978 - https://www.youtube.com/watch?v=sm3ipMFoHoE&list=PL7uw9vTkqQwp1s4oh6TRd8qm6StglCkUd Tropical Events (testing tropical namelist options) Irene - https://www.youtube.com/watch?v=GNhYYvrQrPc&t=252s Bob - https://www.youtube.com/watch?v=FmbhsaFaC4k&list=PL7uw9vTkqQwrX5nzMSy4qFQbTqU9WCqha&index=1 Gloria - https://www.youtube.com/watch?v=60wBgKfBrHg&list=PL7uw9vTkqQwrX5nzMSy4qFQbTqU9WCqha&index=3 Freezing Rain Events (testing p-type algorithm): Dec, 2008 - https://www.youtube.com/watch?v=sNkfVaudVZ8&list=PL7uw9vTkqQwrfqeW1diEdDmR_wGGvgjZC Does anyone have any suggestions for my next 10 simulations? I'm looking for unique events. I'd like to try Hurricane Carol (1954 - test the reliability of ERA20c prior to 1985) and the Blizzard of 2013 next. Dec. 9, 2005 is really interesting too. Before any recommendation(s), I can run simulations with reasonable accuracy from 1978+ (so far) w/ERA5 and ERA20c reanalysis data.

So about those armored millipedes/centipedes... Is there a way to get rid of or prevent them from entering my property? I'm finding them everywhere around my apartment. They creep me out. Something keeps eating my plants too . I had to move them inside.

Glad you guys enjoyed it! There were a couple things that stood out to me: 1) The visibility observations for the RI station near Providence... The station dropped to 1 mile visibility at ~1pm on Feb 6th, then down to ~0.1 miles visibility for ~12 hours afterwards lol. 2) Wind gust >70mph in Boston. This was plotted twice... These observations were taken +-5 minutes from the simulation time stamp. 3) Several hours of mix/rain from Providence to Boston. I don't believe this occurred. For the ERA20C dataset, I can run simulations from 1900 to the current day. Some datasets go back <1900, but they likely won't perform well. I plan on doing this weekly for different events in and outside of the US out of curiosity. We'll see how the older simulations pan out, but the 78' simulation evaluated relatively well.

Offer me some feedback, please! I ran WRF (+ some additional utilities) from 1978-02-04 to 1978-02-09 (for sh!ts and giggles) and created plots for two different grids (parent domain - d01 and 1 nested domain - d02). I added plots of temperature, wind speed, wind gusts, visibility, precipitation type, etc... and superimposed observations when applicable. The still images were looped to create a video and I added audio to, "set the mood." Since I ran a simulation on a historical blizzard, I figured the board would find it interesting. Admittedly, my video editing skills are poor...

This methodology looks reasonable: https://carnotcycle.wordpress.com/2012/08/04/how-to-convert-relative-humidity-to-absolute-humidity/ You won't even need to convert your temperature (C) or relative humidity (%)... Just plug n' chug. To get kg/m3 from g/m3, just multiple g/m3 by 1kg/1000g == 1/1000.

I agree! It's a very good question. Now that it's been a few years, I'd rank it like this; FORTRAN, NCAR Command Language (NCL; which is no longer being updated), and Python. I develop script pretty often in NCL and Python for statistical and graphical reasons. FORTRAN's important because most, if not all, Numerical Weather Prediction programs are coded in FORTRAN. If anyone's curious, the code to run NWP (for the Weather Research and Forecast model specifically) looks like this: https://github.com/wrf-model/WRF/tree/master/phys - Select any .F file in that repository. For Python, the script will look like this: https://wrf-python.readthedocs.io/en/latest/plot.html (I mainly plot images with Python). NCL's pretty similar to this, but faster when working with .nc/.grb/etc... files though. In all honesty, if you learn one language, you obtain a basic understanding of them all.

No. Public websites keep snowfall calculations simple because 1) scripting Cobb's/Dube's algorithm is too computationally expensive 2) in the end, it doesn't matter which algorithm you utilize for forecasts or 3) they can't write the script. Besides what I wrote above regarding positive snow depth, 10:1 ratios is easier to compute: a) If they use a precipitation type algorithm (csnowsfc is boolean wrt snow-> 1==snow, 0==not snow): snowfall = csnowsfc*LWE*10. For snowfall, precipitation type algorithms perform well... Diagnosing mixed precipitation type is problematic though. Here's an article that provides results using "observed" data: https://journals.ametsoc.org/view/journals/apme/55/8/jamc-d-16-0044.1.xm. Go to Table 1. b) If they use microphysics scheme output (SR is continuous -> from 1==all frozen precipitation to 0==no frozen precipitation): SR*LWE*10. You occasionally see websites that state, "this product may include sleet..." They likely used this methodology since SR is a function of what a microphysics scheme diagnoses (mass/concentration of rain, ice, snow, etc...). If it diagnoses graupel, sleet, and hail, SR will include graupel, sleet, and hail as well. It depends on the microphysics scheme. Most modeling systems use Thompson's microphysics schemes which means rain, ice (sleet), snow, and graupel are included in SR. Websites should be more public about their methodologies. Some are vague as He!!. Another pet-peeve of mine, Pivotal provides Kuchera's algorithm for PAID members. Why? Nobody published a paper on its accuracy and I'm sure other SLR algorithms perform better.

That's a good question. The algorithm takes into account "gauge losses" due to melting, compaction, etc... caused by the land-air interface. If land isn't categorized by the model at a given grid point, 0/NaN values will be produced by the algorithm. That's why there are 0/NaN values along the coastline. The (very course) GFS categorizes those grid points/locations as "water." The website post-processes those maps from the GFS by, snowfall(t)=abs(snow_depth(t)-snow_depth(t-1))+snowfall(t-1) - t==time. snow_depth is calculated within the model though. For the coding/scripting weenies: https://github.com/wrf-model/WRF/blob/f311cd5e136631ebf3ebaa02b4b7be3816ed171f/phys/module_sf_noah_seaice.F [via WRF] - search for "SNOWH." I'm assuming the GFS uses the same subroutine. For the non coding/scripting weenies: https://www.wcc.nrcs.usda.gov/ftpref/wntsc/H&H/snow/AndersonSnow17.pdf - The "new snowfall" calculation is denoted by equations 2.a though 3... It's different than what most people are accustomed to on this site. It calculates snowfall by diagnosing snow density.

Thanks for the feedback, everyone! Today is NC + 2 for me. I'm still getting everything setup (furniture/internet), but I'm beginning to get a sense of the area. I live in a nice community surrounded by Massachusetts' equivalent of Route 1 on steroids (15/501). The only sketchy area I've seen so far is along 15/501 near Walmart, otherwise, not bad! Obviously, NC has a much different feel... Even the dirt looks different than in NE (more clay?).

I'm not too big into the NHL. I rank major league sports as; NFL>>>MLB~NHL>>NBA>MLS. The Pat's are my favorite team followed by the Red Sox, Bruins, then the Celtics. I mostly watch the B's and the C's during the playoffs with the occasional regular season game. Fun fact: the first Pat's game I remember is the AFC div. game vs. the Raiders. I have no memory of the Pat's having a losing record... Super impressive! Noted. I won't stick any of my appendages inside any dark, empty void I plan on moving back to MA after a year of work for the EPA. They want me to get the "EPA experience." After that, I can work remotely from anywhere I'd like. My time in NC will be ironic. I got my degree in atmospheric science because I love winter weather. My most fond memories involve snowstorms. To name one, during the blizzard of 05' (Jan 22-24), I remember seeing cars slowly becoming engulfed by snow w/near hurricane force wind. For nearly 12 hours, I remember pure whiteout conditions. When the event ended, I remember going outside and thinking to myself, "I feel like I'm on Hoth." In some places, the snow was taller than I. To this day, nothing compares... Oh! I was ~14 at the time so I was probably 4' 6'' or something like that... So nostalgic! I created snowfall maps for fun. You can view them here (only for the NE US) if you'd like: https://github.com/msw17002/Historical-Precipitation-Analysis/tree/main/Northeast_Snowfall Some new terminology here: Palmetto bugs and noseeums. The worse I run into are horseflies. They're super annoying! I was never bitten by one thankfully. So Magpie meant casual waving! I thought he meant something different lol. Thanks for clarifying that! I look forward to any winter or severe weather event! I'm severe weather deprived... I usually get nothing significant (severe) where I live now. The accents will be fun! I always like listening to different dialects. Oh, the anticipation!

This is great! My mother lived in SC for a few years and her fear of the local wildlife transcended onto me. I exaggerated a little bit for a laugh, but there's sincerity with what I wrote. From what I read online (probably not be reliable), spiders like the Brown Recluse and Black Widow are relatively common throughout NC. I presume this isn't entirely true or nothing to be worried about? I'm looking forward to experiencing the culture change! I imagine there's a pretty big difference b/n people from the NE and the SE in general. The only problem I foresee is that I don't barbecue and I don't watch college sports