Thoughts on accumulation = snowfall rate - melting rate and key variables involved...

[RU848789]

Folks - fyi, I'm not a meteorologist, but I probably know more about heat transfer than 99.99% of the population, since you learn that kind of stuff when you get a PhD in chemical eng'g.  So listen up, lol.  Anyway, we see flawed discussions about this every year, once we get to March and I'm going to try to add some better science to the discussion - I can't say that I have it all exactly correct, but I know it's directionally correct.  Fundamentally, it's pretty simple: accumulation = snowfall rate - snow melt rate, so as long as the snow is falling at a rate that exceeds the melting rate, snow will accumulate, even at midday and even on blacktop surfaces.  In practice, though, it's pretty complicated, as the key question is, what,exactly are the melting rates as a function of ground temperature, air temperature, and insolation input to those temperatures, via radiation absorbed (which is not a constant, but varies throughout the day with sun angle and even cloud deck thickness)?  I've never seen melting rates for these aggregated variables published anywhere for, say, grass and asphalt as a function of time of year (and latitude) and time of day, but that's what we'd really need to know, for sure, what snowfall rate we need to have accumulating snow. 

 

 

In addition, there is another key factor that almost everyone overlooks.  Once there is a snow layer on the ground, the melting rate decreases, as that snow layer, by definition, is at 32F, so the melting of falling snow from "the ground" below it is no longer an issue, as that ground will be in equilibrium with the snow, at 32F.  Furthermore, once there is a snow layer on the ground, the impact of the indirect sunlight greatly decreases, given the much higher reflectance (albedo) of snow vs. the ground: for example snow reflects 85-90% of insolation, whereas asphalt reflects only 5-10%, with grass/trees somewhere in-between.  So, the biggest key in a spring snowstorm, is often getting a period of heavy snowfall that exceeds the melting rate, allowing accumulation - once that happens, melting still occurs, but it is less, such that accumulation will increase for the same snowfall rate.  The huge difference in reflectance is why one can see several inches of snow on the grass/cars little to no accumulation on the asphalt during the day.  Melting from air temps above 32F for the top surface of the snow is still an issue. 

 

 

Bottom line is that if we don't get some sustained, heavy snowfall rates, it'll be like 3/16, where it snowed lightly to moderately most of the day and we only saw accumulation on grassy surfaces during a couple of short periods of heavy snow, but that 1/4-1/2" of snow still melted quickly from above freezing air temps once the snowfall rates fell below about a moderate level.  For storms like 4/6/82, there was sustained heavy snow for hours during the day and we got accumulations, even on roads, once we overcame the melting rate and once that occurred and we still had heavy snow falling, the melting rate decreased and the accumulation rate increased and we got crushed.  It's sounding like this one will be more like the first case, where we won't have sustained heavy snowfall during the day and we'll just see some minor/modest accumulations on non-paved surfaces.  But we can hope...

As an aside, has anyone ever published snowfall melting rates at various temps above freezing at various times during the day (sun angle, really) for various surfaces at various latitudes?  When I was 23, I probably could've written and solved the equations (although maybe not, as this is truly an unsteady state, non-linear system, which is very difficult to solve, analytically, vs. numerically) - not now, lol, as I've forgotten most of that stuff.  Anyone?  Sounds like a pretty cool dissertation topic...

[RU848789]

Well, we got to see a lot of the above in action today.  Been at work all day and evening and don't have time to find the data, but I wonder how much snow fell (liquid equivalent) in areas that had no accumulation, like, say, New Brunswick and how much snow fell in areas like AC or Cream Ridge, where a few inches of snow fell - would be interesting to see the hourly precip rates, snowfall accumulation rates and, by difference, the melting rates in these cases (or more cases) to try to assess the melting rate and its contributing variables, during the storm.  Still curious if any of the pros/NWS have data like this, in general, or if it gets collected for specific storms like this one.  Also, I must say, I'm a little disappointed that this post, which I think discusses an interesting meteorological topic, has had no responses yet...

[Isotherm]

Great posts RU.

 

I'm in interior Monmouth County, and recorded 0.65" of precipitation yesterday. For a time midday, around 11am-1pm, the grass turned white in spots, and I may have accumulated 0.1" or 0.2" on the snow board briefly, but it quickly disappeared by 2pm once the precipitation rate decreased.

 

The threshold for accum on grass seemed to be 0.09" of precip per hour, but this was after my temp had fallen to below 34.5F, and bottomed out at 33.7F temporarily around 12 noon.

 

My precip rate at 9am: 0.05"/hr

10am: 0.06"/hr

11am: 0.09"/hr

12 noon: 0.07"/hr (the 0.1" melted off)

1pm: 0.10"/hr (peaked at about 0.2" on the snow board)

2pm: 0.07"/hr (melted off)

3pm: 0.04"/hr (warmed to 36F and nothing sticking to grass)

 

So for my area, if we could have kept up a rate of 0.09-.10"/hr with temps around 34F, snow would have accumulated. This was when the 25 dbz bands reached my area. Problem is the storm wasn't organized enough to produce a LONG LASTING deformation band as seen in major late season snowfalls like march 1958, 1956, and April 1982, and even April 2003 featured hvy rates for several hours.

 

One hour of hvy rates is enough to cool the boundary layer and allow the accum rate to outweigh melting rates, but if we then return to light/mod rates, that "work" done will go to waste via more melting.

 

At this time of year, especially in daylight hours with marginal temperatures, we need sustained rates of 0.09-0.10"/hr or > sustained over several hours in order to accumulate. Simply put it just didn't snow hard enough for long enough time, though I recorded 0.65" of precip total here due to moderate precip the entire day. If this event occured a month ago, it probably would have been a significant snowfall IMBY.

[jm1220]

Great posts RU.

 

I'm in interior Monmouth County, and recorded 0.65" of precipitation yesterday. For a time midday, around 11am-1pm, the grass turned white in spots, and I may have accumulated 0.1" or 0.2" on the snow board briefly, but it quickly disappeared by 2pm once the precipitation rate decreased.

 

The threshold for accum on grass seemed to be 0.09" of precip per hour, but this was after my temp had fallen to below 34.5F, and bottomed out at 33.7F temporarily around 12 noon.

 

My precip rate at 9am: 0.05"/hr

10am: 0.06"/hr

11am: 0.09"/hr

12 noon: 0.07"/hr (the 0.1" melted off)

1pm: 0.10"/hr (peaked at about 0.2" on the snow board)

2pm: 0.07"/hr (melted off)

3pm: 0.04"/hr (warmed to 36F and nothing sticking to grass)

 

So for my area, if we could have kept up a rate of 0.09-.10"/hr with temps around 34F, snow would have accumulated. This was when the 25 dbz bands reached my area. Problem is the storm wasn't organized enough to produce a LONG LASTING deformation band as seen in major late season snowfalls like march 1958, 1956, and April 1982, and even April 2003 featured hvy rates for several hours.

 

One hour of hvy rates is enough to cool the boundary layer and allow the accum rate to outweigh melting rates, but if we then return to light/mod rates, that "work" done will go to waste via more melting.

 

At this time of year, especially in daylight hours with marginal temperatures, we need sustained rates of 0.09-0.10"/hr or > sustained over several hours in order to accumulate. Simply put it just didn't snow hard enough for long enough time, though I recorded 0.65" of precip total here due to moderate precip the entire day. If this event occured a month ago, it probably would have been a significant snowfall IMBY.

You really needed to be in the heart of the deformation and CCB to have a real shot at getting much. That's why even when people were mentioning 0.50"+ qpfs on the models, it was still the same as a miss even if those amounts verify, because that amount fell throughout the day and intensity was just never enough to even keep it all snow. You really need a walloping for it to count this time of year.

[NorthShoreWx]

Great posts by RU.  I don't have any math to add to the discussion unfortunately, but I can add subjective interpretation of my personal observations over the years

 

Melting snow does cool the interface with the ground, although there is usually more heat lurking just below outside of the heart of winter at this latitude.  In the absence of WAA, I have found that heavier rates of melting wet snow will cool the boundary to within a cats whisker of freezing just about every time.  It usually won't start accumulating until the temp drops below ~34.  Once that air temp is close to freezing and heavy wet snow is accumulating, you can see some fairly good S/W ratios as long as it continues.  Warm it up a little, or mix a little rain, that's when the ratios drop.

 

Assuming you can't get good ratios with "wet snow" (based on the surface air temperature) is one of the many misconceptions too often proclaimed around here.

 

Another misconception I see a lot around here is that a forecasted high temperature of say, 42, is inconsistent with a forecast for snowfall.  I've seen highs in the 40s early in the day many times only to drop quickly with the onset of snowfall.  If the column is cold and there is no major WAA at the surface, it's all about the precip rates.  And it does not require that we be parked under a CCB to get accumulating snow (it does help for the rates, but it is not the only way).

[mweisenfeld]

Positing from a total lack of knowledge here, but wouldn't you also need to consider humidity?  Melting takes two forms, transfer to liquid that flows down to the ground, and evaporation (similar to virga but from the ground).

[NorthShoreWx]

Positing from a total lack of knowledge here, but wouldn't you also need to consider humidity?  Melting takes two forms, transfer to liquid that flows down to the ground, and evaporation (similar to virga but from the ground).

 

Yes initially, but once the air is nearly saturated evaporation steps to the rear of the bus in terms of responsibility for heat transfer.

[Weathergun]

At least 2-3" per/hr snowfall rates, three hours straight, is what I want to see for significant accumulation on the paved surfaces, with a late March/early April sun angle in NYC. That comes with UVVs higher than 15. Also helps to have less of a marine influence, with north winds. And wet-bulb temps starting out below freezing, for evaporative cooling.

[NorthShoreWx]

At least 2-3" per/hr snowfall rates, three hours straight, is what I want to see for significant accumulation on the paved surfaces, with a late March/early April sun angle in NYC. That comes with UVVs higher than 15. Also helps to have less of a marine influence, with north winds. And wet-bulb temps starting out below freezing, for evaporative cooling.

I think that's overkill, but I'll take it.

[Isotherm]

I think that's overkill, but I'll take it.

 

 

Agree.

 

For the April 7th, 2003 event, CNJ was under 25 dbz echos for 3-4 hours during the middle of the day. That's moderate to hvy snow but it's probably 1-2"/hr rates, not 2"+/hr.

 

Additionally, the cold air source was phenomenal, with a 1040+mb sfc high to the north funneling sufficient boundary layer air and lowering the wet bulb.

 

http://www.raymondcmartinjr.com/weather/2003/07-Apr-03-FortDixDopplerRadarImagery.html

 

Radar imagery per Ray's site. Again, not extremely hvy precip, but moderate intensity for several hours. Solid banding with sfc temps near 32F on a N/NE wind yielded 8" IMBY.

[bluewave]

It snowed hard enough on 4-5-06 to drop the temperatures from 42 to 32 in a short amount of time.

You can see form the video that the snow was able to stick on some of the paved surfaces

other the the center of the streets.

 

One of my favorite blizzards of all time was April 6th, 1982. It was remarkable to see true blizzard

conditions in early April with temperatures falling through the 20's and lightning.

 

http://www.wunderground.com/history/airport/KJFK/2006/4/5/DailyHistory.html?req_city=NA&req_state=NA&req_statename=NA

 

 

 

 

 

[RU848789]

Great posts RU.

 

I'm in interior Monmouth County, and recorded 0.65" of precipitation yesterday. For a time midday, around 11am-1pm, the grass turned white in spots, and I may have accumulated 0.1" or 0.2" on the snow board briefly, but it quickly disappeared by 2pm once the precipitation rate decreased.

 

The threshold for accum on grass seemed to be 0.09" of precip per hour, but this was after my temp had fallen to below 34.5F, and bottomed out at 33.7F temporarily around 12 noon.

 

My precip rate at 9am: 0.05"/hr

10am: 0.06"/hr

11am: 0.09"/hr

12 noon: 0.07"/hr (the 0.1" melted off)

1pm: 0.10"/hr (peaked at about 0.2" on the snow board)

2pm: 0.07"/hr (melted off)

3pm: 0.04"/hr (warmed to 36F and nothing sticking to grass)

 

So for my area, if we could have kept up a rate of 0.09-.10"/hr with temps around 34F, snow would have accumulated. This was when the 25 dbz bands reached my area. Problem is the storm wasn't organized enough to produce a LONG LASTING deformation band as seen in major late season snowfalls like march 1958, 1956, and April 1982, and even April 2003 featured hvy rates for several hours.

 

One hour of hvy rates is enough to cool the boundary layer and allow the accum rate to outweigh melting rates, but if we then return to light/mod rates, that "work" done will go to waste via more melting.

 

At this time of year, especially in daylight hours with marginal temperatures, we need sustained rates of 0.09-0.10"/hr or > sustained over several hours in order to accumulate. Simply put it just didn't snow hard enough for long enough time, though I recorded 0.65" of precip total here due to moderate precip the entire day. If this event occured a month ago, it probably would have been a significant snowfall IMBY.

Awesome data Isotherm.  I was kind of thinking it might take an inch per hour snowfall to start accumulating on grass during the midday hours with temps in the mid-30s and you confirmed that.  I'd guess it might take 1.5" per hour snowfall rates to get accumulation on paved surfaces - we know both are obviously possible, based on what has happened in many past storms (and in SE PA/VA/MD/South Jersey yesterday).  Then, once there's a solid snow layer, I'm guessing the decreased melting rate from having that layer might be 0.7-0.8" per hour due almost solely to air temps (since snow reflects incoming radiation so well).  Alas, we just never got there in Central Jersey.  Still think this would be really cool data to have - my guess is it wouldn't be too hard to simply model it with known conditions and known melting rates for the various sub-elements leading to melting.