This is a full play-by-play of the eclipse with some technical details and minor analysis. Hopefully someone out there reads this and gets an idea of what maybe to do and not to do with their cameras during an eclipse. Click photos for full size ones.

I had identified the totality of the solar eclipse of Monday, August 21, 2017 as one I wanted to see years ago. It was the first one that I had access to see within my lifetime with plenty of options for travel within the United States to see it. I had planned on scheduling a big trip over a year out to see it, but the discovery of brain cancer and complications from it throughout 2016 led me to believe there was a reasonable chance of not being alive to see it. So, I waited until later December, 2016 to decide on a trip when it had become clearer that the tumor, or complications from it, would not kill me in the short term. I ultimately chose the Charleston, SC area, specifically Isle of Palms to the east/northeast as the viewing location. My other options seemed to be western KY or Tennessee, but neither provided a lot of options for entertaining my two kids. Given that my kids love the beach, Isle of Palms seemed like a no-brainer choice for a viewing location even though it would literally be the last place to see it along its path because the parts of SC further to the northeast within totality on the map along the coastline are uninhabited.

Sometime between then and the eclipse, the marketing and excitement within the general public for the eclipse began and the resort I was staying at, Wild Dunes, began a marketing campaign of “Dunes Go Dark”. Eclipse fever was in full effect when we arrived Friday later in the afternoon. For the week before the eclipse, the forecast was for partly cloudy skies with a 20% chance of showers on eclipse day. That was true as of that Saturday morning, but by Saturday afternoon the forecast for that Monday had changed to mostly cloudy skies with a 50% chance of showers. That forecast stayed the same through that Monday, with only the chances of showers varying from between 30% and 50%. Ultimately, the final forecast from forecasters was that skies would be cloudy and showers would be along the coastline for the morning and early to mid-afternoon before redeveloping inland sometime in the afternoon. I had previously decided I would drive to the northwest corner of South Carolina or far western North Carolina if the weather looked like it could be bad, but I decided to risk it and stay put on the beach for the eclipse.

At the time I booked the trip, I purchased a pair of plastic eclipse glasses and four of the cheaper paper eclipse glasses commonly seen. I also purchased an 8’x8″ solar filter sheet of the same material seen in the paper glasses to make a cheap filter for my camera for the partial parts of the eclipse. I definitely wanted some photos of the eclipse in all of its phases. My primary camera and lens combination is a Nikon D7100 with an 18-200mm lens. I had previously taught myself manual modes of DSLRs when I first got a DSLR in 2009, but I lost what I had learned and largely just became a “dad with a camera” using automatic mode almost exclusively.

A couple of months prior to the trip, I made my filters for my camera and took some test shots of the sun. My original plan was to try to do a time-lapse of the entire eclipse, if the skies were relatively clear, by using the filter during the partial phases of the eclipse and no filter during totality. I quickly discovered that a full time-lapse was not even close to possible because the sun moved too quickly across the sky. Even at 50 or 70mm, the sun moved too quick across the frame to be able to make a fairly seamless time-lapse. The test shots of the sun turned out well through the filter, though, and I jotted down some manual settings that I felt may help if there were light clouds or haze even though automatic mode seemed to work pretty well.

I then did some test shots of the nearly full moon at 200mm using some of the settings listed here:

 

The Nikon guide on photographing an eclipse helped too:

What I found was that all the photos of the moon looked relatively similar, but did not reveal any of the finer details of the moon and its craters, as I felt they should. Historically, my camera has always taken shit pictures of the moon no matter what I tried and this is the reason there was no way I was going to rely on automatic mode for the eclipse. It was only when I jacked up the ISO to 1600 in this series of tests that the details became a little clearer. I ultimately decided on an ISO of 1600, an f/number of 22 and a shutter speed of 1/200 as what I would attempt to photograph the eclipse with and those are the settings that were used to take this photo of the moon. The moon turned out significantly darker than it was to the eyes in this test, but it seemed like the best the camera would do:

According to the chart, those settings would give me some of the corona, which is all I wanted to capture as a memory, but really had no clue as to what I was going to get or if the photos were even going to look good at all. I also decided that since I was not going to be able to do a full time-lapse that I may take the camera off of my tripod right before totality so that I could also snap some quick pictures of the way the sky looked around the beach, as I figured this may look pretty cool too. But, I was bringing my old D70 for that, or as a backup camera in case my primary one somehow failed.

The day before the eclipse, I went on a fossil hunting trip in Summerville, SC and pulled out my old Nikon D70 to take some photos. The D70 game me a “CHA” error after snapping a photograph and turning the camera off and on allowed for another photo to be taken before the error occurred again. The same thing happened a third time. I didn’t try to take any more photos and when I got home, the photos were corrupted, and a quick internet search led me to believe that the CHA error is a result of a corrupt compact flash card or the camera having issues writing to it. So, my backup camera had failed.

Monday, August 21 arrived and the weather was mostly shit for viewing an eclipse. Skies were almost completely cloudy as a whole, a result of partly cloudy skies at the low, mid and high levels. There were even some sprinkles later in the morning, but a look at the radar showed the near-stationary thunderstorm bands to be forming about 20 miles inland and about 20 miles offshore, leaving us in the middle to be largely rain-free. The radar couldn’t be completely trusted, however, because the Charleston radar had failed the day before due to a lightning strike and wouldn’t be fixed until at least Tuesday, and the radar I was looking at was from far-away Columbia, SC. Regardless, my wife and I both remarked that it was great beach weather, if it wasn’t eclipse day, as the clouds made it feel nice without the sun beating down on us.

By partial eclipse start time, the sun had become mostly visible through the high clouds and I got some photos on a five second timer. They weren’t nearly as good looking as the test shots I had taken back at my house in Baltimore in clear skies, but they don’t look terrible either. Also, the videos didn’t show the transit of the moon nearly as well as I had hoped they would.

Here’s one just as the eclipse started:

Here’s another that seems overexposed after I manipulated the f/number and shutter speed in an attempt to make it look better through the clouds:

The last video shows the sun going behind a cloud deck and the sun continued to not be visible at all back behind cloud decks at all three levels:

 

It was especially the mid-level cloud deck that was the problem. This lasted for 45 minutes until about 10 minutes before totality. At this time, the low-level clouds started to dissipate and the mid-level cloud deck started to move away resulting in the sun being visible through the high cloud layer. People began cheering as the sun was now visible and nearly totally obscured by the moon. The low incoming solar radiation had made the sky look weird, too. It wasn’t like the evening; it was just different and a little eerie as pale sunlight came down from above. In the week before the eclipse, I had read a number of articles talking about animals acting unusual, like dolphins coming up to the surface of the ocean, starting about 15 minutes before totality. I can totally see why a more intelligent animal, like a dolphin, would come up to check out what was going on given the unusual light. This Gull’s behavior was not abnormal, however, as it was flying low looking for food in the low light that the crowd may have dropped on the beach:

I took my camera off the tripod. At this time, I snapped this photo without a filter and the time index indicates it was nine minutes before totality. Playing with the .nef file revealed just how many high clouds were remaining in place for the approaching totality:

Within a minute, photos without a filter were too bright because the clouds low and mid-level clouds continued to dissipate. I then snapped this photo with the filter. It didn’t turn out as well:

I kept briefly looking up at the sun, hoping that I didn’t blind myself. The photo through the filter overexposes the sun making the crescent look bigger, as I could see, and the previous photo showed, that it was just a small sliver at the time. With only a minute or so to go to totality, some younger women a few beach blankets towards the ocean started playing “Total Eclipse of the Heart” by Bonnie Tyler:

Then, totality came and the dunes went dark. The sun’s corona was clearly visible to the naked eye, despite the high cloud cover and the crowd continued to cheer. I had switched to the manual settings that I had wanted and snapped a quick picture as it entered totality. When I looked at it later, I wasn’t sure what was going on in the bottom left quadrant of the picture in the reddish area, but research leads me to believe the camera was picking up the chromosphere. As to why it’s blurry, it could easily be caused by the clouds, it could easily be lens shake from me, it could be features of the chromosphere itself, it could be blurry bailey’s beads, or some combination of all four:

Subsequent photos clearly show red solar prominences extending out from the sun:

Both the chromosphere and solar prominences being visible in the photos were completely unexpected to me and these were not visible to the naked eye. It’s curious that neither my wife or I saw them, yet they are clear in the photos and that they are clear despite the high cloud cover. It’s also curious that Mr. Eclipse’s chart believes a much quicker (in fact, 10x quicker) shutter speed is necessary to capture prominences. Did the high clouds somehow bring them out? But also, the corona was a little less expansive in the photo than I was expecting. Surely the clouds were an inhibiting factor for that?

I then switched to automatic mode, just to make sure those pictures weren’t better and snapped several. They were shit, as expected, and this is by far the best one I got:

I then very quickly snapped a few photos of the way the sky looked towards the ocean, which was similar to sunset:

The bright spot in the sky in this photo is not a star, but someone’s drone filming totality:

But then, the color in the sky was lost as we entered the middle of totality:

At this time, my daughter noted what was likely Venus shining brightly in the sky. But, we couldn’t see any other stars given the cloud cover. We could also see flashes of lightning from the thunderstorm band to the north. At first, I thought it might have been people’s flashes on their camera, but it was definitely lightning without the thunder. This is what some people call nighttime heat lightning, which is a misnomer. I switched back to photographing the eclipse on manual and snappedtwo more photos.

I then took some wider angle shots on the same manual setting in the hopes of picking up Venus and perhaps some of the clouds, but that didn’t happen:

What’s noteworthy, though, is that the solar prominences are still visible, indicating that their visibility was definitely not a function of my lens extending to 200mm, as these shots were taken around 32mm.

A closer look at the eclipse at 32mm, still showing the prominences:

I then went back to 200mm as the eclipse was exiting totality and got a series of shots, ending in the diamond ring effect:

Every year or two I take a photo that becomes my new “favorite photo” that I’ve taken and one of the final photos of the diamond ring effect, with the “diamond” flanked by solar prominences on each side is my new favorite photo. My imagination tells me it’s a diamond surrounded by rubies. I like that it also highlights the fact that clouds were there, as that was part of the eclipse experience. I made a quick widescreen wallpaper out of it with some minor adjustments:

Here’s an animated .gif of all the photos taken of totality at 200mm. I tried to line them up, but it’s not perfect:

The sun came almost fully out shortly after totality and I was able to snap some pictures that were closer to the original test shots of the sun at home:

Here’s a last video as the sun was lowering in the sky:

With about 20 minutes of partial eclipse left, the thunderstorms started migrating south towards us slightly, as they reformed on their own outflow boundary, and threw some more high and mid level clouds over the sun, obscuring it. I could tell that was it for the eclipse, so I turned off my camera and started packing everything up to head back to the resort and make some backup copies of all of the photos.

Analysis of the photos taken around the beach indicate to me a possible reason why the low level clouds dissipated shortly before the eclipse and that is because the eclipse itself was occurring. I believe the significantly reduced solar radiation prior to totality may have broken the convection currents happening at the time causing the low level cumulus clouds to dissipate. From a general perspective, a daytime convective current at the beach looks like this:

At the cloud level, lower level cumulus clouds form as a result of the sun heating the surface of the earth. The air rises, causing the cloud to form, then the air descends on either side of the cloud in return flow. With the solar output reduced, the surface of the earth wasn’t warming, thus the convection currents may have been broken.

I’ve circled and drawn some arrows on one of the previous photos on a set of clouds that I believe shows this. While the clouds’ initial convective structure is still intact, these clouds out over the ocean are clearly in a dissipating stage, as they look tilted, jagged and weak:

Photos looking east/northeast still show some thicker lower level clouds and even a sprinkle out over the ocean, so it doesn’t fully explain the dissipation over my exact locality, and it doesn’t explain the mid-level clouds luckily departing, either.

Taking the photos and looking at them later has been fun. While I did a little bit of homework prior to the eclipse, I would say it was still about 95% luck that the photos turned out as well as they did with the low-level clouds dissipating and mid-level clouds departing and the chromosphere and prominences unexpectedly turning up in the photos. If you’re a dad or mom with a camera, amateur or novice DSLR user and I had to summarize some bullet points for photographing an eclipse, they would be:

  • Make sure your camera is set to write raw files (.nef files for Nikons) or jpg and raw files simultaneously. I had gotten away from this practice, but made sure raw output was being written prior to the eclipse.
  • Make your own cheap solar filter if you want, or don’t. The pictures are only a little better than taking a camera phone and snapping a picture through eclipse glasses. The photos are impersonal and difficult to take if there’s any clouds.
  • The time-lapses would be cool, though, if the transit of the moon can be shown clearly. Obviously, get a tripod if you’re going to do this. My interval was five seconds and I don’t think there’s such a thing as taking too many photos in a time lapse, so the interval could have been shorter. It’s just dependent on the total number of photos the camera allows in a sequence and how long the battery is going to last before you have to switch it.
  • Do your homework and find the best possible setting for taking a photo of the moon with your DSLR. Hopefully this translates well to the low light of eclipse totality.
  • I totally forgot to test out bracketing of shutter speeds or set this up before the eclipse. This is something that should be done.
  • Write all of your settings that you decided on down on a piece of a paper so that you don’t lose track of what you’re doing. There’s only several minutes, at best, to see a total eclipse and capture it and you’ll want to look at it with your own eyes too.
  • Get yourself a pair of plastic eclipse glasses, not the paper or cardboard kind. It will be much easier to work with them in the partial eclipse phases.
  • If your camera lens doesn’t have the eclipse filter on it, but it’s pointed at the sun on a tripod, cover it with a towel or something so it doesn’t get overly hot.

Before this one, the last total solar eclipse visible in the contiguous United States was in 1979. The next one will be in 2024:

Looking at the map, I can’t be the only one thinking a trip to Niagra Falls is in order? Maybe I’ll get some even better photos when the Falls Go Dark.

***
Here’s a few more photos from the Isle of Palms and Capers Island:

 

 

 

A tornado hit Kent Island in the middle of the night last night. The Mount Holly, NJ National Weather Service declared it an EF2 tornado at ~125mph, but the photos of damage done to Kent Island indicate it is the strongest tornado to strike Maryland since the F4 April, 2002 LaPlata tornado that also went across the bay to be an F3 tornado in Dorchester County, MD.

Photos this morning showed extensive damage to some houses, others shifted off their foundations, cars and boats tossed around, and numerous large trees shredded or uprooted. Everything can be found on social media, but I wanted to comment that this appears to me to be another tornado forming as a result of the sudden friction difference between water and land. It’s widely taught on television during approaching hurricanes that when hurricanes strike land, tornados often form within the rain bands due to the sudden encounter of friction with land. It’s actually the friction difference and tropical systems that still have strong circulation that are later exiting the continent onto the ocean (which happens in some cases) also exhibit the same behavior.

The same appears to be true with “regular” thunderstorms or thunderstorm cells of a non-tropical origin – difference in friction can turn them into full-blown mesoscale low pressure systems with a strong tornado. Our local National Weather Service noted on a now defunct web page that the LaPlata tornado in 2002 was likely initiated by the strong thunderstorm crossing the Potomac River from Virginia. With the Kent Island tornado, the thunderstorm was strong in Prince George’s county (there was a report of a tree down in Glenn Dale), then went through Anne Arundel County, entered the Chesapeake Bay non-tornadic and by the time it was on Kent Island only 4.3 miles later, was obviously a full-blown mesocyclone with a strong tornado. It seems likely to me that the thunderstorm exiting the western shore of Maryland onto the Chesapeake Bay was a factor in initiating and forming the powerful tornado.

It’s fortunate that nobody was killed in this storm.

***

Justin Berk stated it was likely an EF0 or EF1 tornado around 1pm today after photos of damage clearly indicated it was EF2 or higher.

There were indications this was a major tornado (EF3), as exterior walls of houses fully collapsed, but its dependent on the construction of the house and that is a subjective call after the fact. The NWS usually goes conservative in their estimate, so the EF2 rating is not that surprising. Berk later stated Doppler radar estimate of winds was 86mph and that is presumably why he felt it was an EF0 or EF1.

***

Before the NWS assessment came out, a number of outlets stated the maximum wind gust was 68mph. I tracked down this obviously erroneous information and found that it came from a buoy on the western shore near Eastport and that this was the highest wind speed recorded in the local storm reports from the Sterling, VA National Weather Service forecast area. Why news outlets were saying this was the wind speed on Kent Island is beyond me, but it’s further proof that the mainstream media refuses to do its job and warrants being given the name “fake news”.

000
NWUS51 KLWX 240533
LSRLWX

PRELIMINARY LOCAL STORM REPORT
NATIONAL WEATHER SERVICE BALTIMORE MD/WASHINGTON DC
133 AM EDT MON JUL 24 2017

..TIME…   …EVENT…      …CITY LOCATION…     …LAT.LON…
..DATE…   ….MAG….      ..COUNTY LOCATION..ST.. …SOURCE….
..REMARKS..

0124 AM     TSTM WND GST     1 E EASTPORT            38.97N  76.46W
07/24/2017  M68 MPH          ANZ532             MD   BUOY

&&

EVENT NUMBER LWX1702513

$$

DH

A major international news story broke this past week: a mysterious new land mass has risen from the Atlantic ocean overnight in the middle of the Bermuda Triangle. Full of old shipwrecks and whale bones, it’s a very dangerous place to get to because sharks and stingrays constantly circle it. Few are fortunate enough to get to it after a long voyage at sea and many rugged explorers have nearly perished in their attempts. But for those who do make it, untold treasures await.

At least that’s what the mainstream media wants you to believe. Except that it’s a sandbar just off of Cape Point on Hatteras Island near Buxton, NC and arguably part of the point itself.

First, the articles:

Next, what the island really is. As the featured title image above shows, Chad Koczera took a photo with his drone (great shot!) towards the end of May of a sand bar that had formed off of Cape Point on Hatteras Island near Buxton, NC. At some point over the month of June, the story of this “new island” got legs, no doubt at least partly because it’s a great photo. Fortunately for us in this day and age, we have google earth with historical satellite photos so that we can take a look at Cape Point through time and determine if this is indeed a “new island” or not.

In the animated gif below, that starts with a satellite image of the point in February, 2017 and then tracks back in time to February, 1993, one can see the “island” beginning to form this past February. I’ve placed a red line just below the island and all that was needed at this point was for currents to bring in a few more inches of sand for the island to form. As the gif tracks back, you will see that Cape Point is in a constant state of flux with the point sometimes pointing towards the southwest as currents push it that way and as opposed towards the current southeast. More importantly, the point often extends to the red line south of the island as recently as 2014 and 2011. In 2006, the point extended south of the red line and in 2005, a sandbar suspiciously similar to the current sandbar was forming well south of the red line and current sandbar just after the 2004 photo showed virtually no point at all.

It’s not difficult to conclude that the sand in the shallow water of that area is constantly morphing as the currents do with it as they please. In fact, it’s basic earth science that coastlines are in a constant state of change, especially barrier islands (this was ninth grade when I was 14 years old in my school system, but others may be earlier). As for the reason this “island” forming is news, I believe the mainstream media has long-since tricked itself and others into believing coastlines are static as part of catastrophic anthropogenic global warming theory, which states we are all going to die from rising coast lines, among other things, and that without the industrial revolution, coastlines would be static (which has always been laughably stupid). It’s also is telling that if someone takes a nice photograph (in this case of the sandbar from the drone), the media will then build a fantastic story around it (without any sort of payment to the photog, I’m sure).

All of that having been said, many of the articles mention that it is simply a sandbar and that officials believe it could disappear at any time, but that is often at the end of the article after the average reader with the attention span of a 1.5 year old has long since moved on to the next bullshit article. I also have to give credit to the Huffington Post, which specifically reported someone stating the island is not a result of “climate change.” This is shocking for the HP, actually, but perhaps they note this because it’s new beach and not erosion taking place?

But some of the fantastic elements of this story show how bad media has become. The Mirror article can easily be thrown out as straight up tabloid nonsense (even though certainly some will believe it), but other articles probably aren’t thrown out as nonsense in people’s minds and those are fake news. The articles that mention shipwrecks and whale bones on this island are straight up fiction (and yet another embarrassment for fake news BBC), but a small amount of knowledge and research of the OBX shows how this information made its way into the article(s). The Diamond Shoals area of the OBX has long caused shipwrecks in the OBX area and research shows that Cape Point is considered the beginning of the Diamond Shoals area. As for the “whale bones”, the entrance to the Outer Banks by vehicle is known as “Whalebone Junction.” It’s called this because a businessman put a whale skeleton that had washed up south of Oregon Inlet at his gas station there and essentially renamed what was called “the junction” to Whalebone Junction. The whale bones aren’t there anymore.

As for the sharks and stingrays angle other articles take, yes, I’m sure sharks and stingrays are there, as they are everywhere in the Atlantic Ocean, but I doubt there’s any more there than at any other spot and I haven’t heard of any of the shell collectors getting eaten by sharks or stung by stingrays. As for it being a “land mass”, yes, I suppose it’s technically true, but I personally wouldn’t call a temporary sandbar a land mass.

I hate to have to use the term “debunked” because that’s what every global warming alarmist has used for any argument on the skeptic’s side since the mid-2000s and is way overused, but I believe the animated gif I made debunks the new island story. Looking at the satellite imagery and putting on my critical thinking cap, I believe that sandbar is simply a part of the point itself and within a few months the island will indeed “disappear” by joining the point to the north resulting in an elongated north-south oriented point seen in some of the satellite images before the currents take the point towards the southwest again before starting the whole cycle over. Or perhaps the seas will rise up and suddenly swallow us all as the current paradigm of climate science states.

“I got an idea that I wanna share. You don’t like it? So what? I don’t care.” -Rob Base.

Every year, the big cities in the Mid-Atlantic and/or northeast US are treated to hype for a big storm that ultimately busts for those cities and inexorably, the post-storm shit show of finger pointing happens. It took until Tuesday, March 14 this winter, but the same broken record finally played again with the all-too-familiar song where the main refrain wails “we need to be better at communicating uncertainty in the forecast”. In case you were in a cave and missed it, a potential blizzard was set to blast DC to Boston with actual blizzard warnings for the Philadelphia and New York city areas. The storm formed, but formed west of where models and forecasters thought it would resulting in mixing for the major cities with the heavy snow inland.

The post-storm shit show kicked off with an AP article that stated weather forecasters thought on Monday afternoon that the big cities would be spared of large snow accumulations but decided to keep the forecast and blizzard warnings in place to ensure continuity and not confuse the public by flip-flopping the forecast.

Not too long after that article, opinion pieces popped up in various places blasting the National Weather Service for its decision to keep the large snow forecast in place if they thought it wasn’t going to happen. One of the more notable ones came from the Capital Weather Gang (paywalled), complete with an obligatory dig at Trump. Social media, such as twitter, then lit up with every Tom, Dick and Harry injecting their opinion on the subject.

First of all, the general public can go fuck themselves with whatever opinion they have on anything. Their primary concern is what exactly is the size of Donald Trump’s hands. And those are the smarter ones. While I would be considered a member of the general public, I have more of a stake in storms like this because I issue actual forecasts to friends for the Mid-Atlantic for winter storms such as the one being discussed. Thus, the reason for my opinion in this post.

Second, the National Weather Service forecasters have always been and always will be in a no-win situation with these storms. It doesn’t matter if the forecast is even slightly off, or entirely correct, people are going to bitch and moan anyway because bitching and moaning is woven into the very fabric of American society. Plus the danger in going to school, work, driving, etc. was all still there in the heavy sleet and freezing rain instead of heavy snow. What people were actually pissed about was the fact that they couldn’t have a good snowball fight or build a decent snowman prior to going back in the house for hot chocolate on their snow day off.

The general public, with their need for everything reduced to 140 characters or less, isn’t going to care about the uncertainty of a forecast, so fuck them. Nobody who has stumbled across this post is reading it several paragraphs in, only search bots archiving it are. (That’s the reason for posts like this and ones in my prior iterations of blogs – they’re for when humans kill themselves from their own stupidity and a future intelligent species or aliens are able to uncover these writings, they’ll understand I wasn’t part of the problem.) Anyone remotely enlightened about east coast USA weather understands that every coastal low, regardless of forecast, is uncertain in its strength and track. So the constant calls in the weather community for communicating “uncertainty” better are a meaningless circle jerk.

Given the very cold air mass in place (did you notice how cold it was the day after the storm?) and the likelihood that it would stay below freezing for the duration of the storm, I agreed with the National Weather Service for keeping the warnings up. But then, our local National Weather Service forecast office downgraded our winter storm warnings to winter weather advisories despite ongoing heavy frozen precipitation. As I noted here, this was baffling to me and the first time I’ve ever seen this occur. So if I’m understanding the AP article correct, the NWS kept the warnings in place because they didn’t want to confuse the public by flip flopping, then flip flopped the forecast for the Mid-Atlantic mid-storm, potentially confusing the public, despite no need to flip flop, by definition.

Our local NWS is deserving of criticism for this flip flopping in my opinion. Also, as hard of working as the NWS forecasters are, I believe most of the actual professional meteorologists can go fuck themselves too with their post-storm opinions on the forecast of these storms. Most (but not all) professional meteorologists have stood by and said nothing for the past decade as money gets pumped into the great global warming fraud rather than weather. On the exact same date as this blizzard-that-wasn’t back in 1993 was the March, 1993 Superstorm. People who paid attention to the weather at the time would know this was a watershed moment for meteorology. Models predicted the massive storm over three days in advance and all of the professional meteorologists patted themselves on the back for the model performance. Now exactly 24 years later, the models have crazy fast computers to work with, very tight grids, and a crap ton of data to work with and the performance in these types of storms for real world forecasts is exactly the same as it was in the 90s, on average.

Politicians, too, can go fuck themselves. Why the Post even wasted typing clicks on Jabba the Hutt, I mean Chris Christie’s opinion on the subject is beyond me (“thought leader” ? LOL!). As for Trump, I don’t have that much of an issue with what he said, because he’s right – these storms usually turn out to not be as bad as they are forecasted or hyped for the big cities. That’s the reason the weather community has this discussion all the fucking time. Remember when Major Hurricane Matthew caused catastrophic damage on the Florida peninsula and Orlando last October? Me neither, because the center stayed offshore the entire time despite hype that it could go inland. And incredibly, NOAA plans to update the GFS model in May with one that is known to be inferior for forecasting hurricanes.

Jabba the Hutt has “had his fill” of the National Weather Service. I think he’s had his fill, and more, in general.

The real people here who require an accurate forecast are emergency managers and planners. What is their opinion of the situation? I have no idea thanks to a media who for eight years said “look how cute Obama is using a selfie stick” and now says “Trump has allowed the Russians to overthrow the United States”. The media refusing to do its job for the past 20 years and the dumpster fire that social media has turned into means that the voices of the people whose opinion counts aren’t heard.

Finally, a number of professional meteorologists, as well as hobbyists, who were basing their forecast on the euro model, claimed their forecast and the euro model forecast verified. This is idiotic, to say the least. It was only starting with the 12z March 13 euro when it became very clear that the euro had the storm hugging the immediate coastline. That was only about eight hours away from the start of the storm by the time the model was finished around 2pm eastern. Emergency managers and others need days of lead time to prepare cities – if there was no blizzard forecast and then one forecasted to occur within 8 hours, an even more massive shit storm would have ensued. This is exactly what happened in the epic fail storm of January 25, 2000, except many people had already gone to bed when the warnings were issued for heavy snow prior to them waking up. This article here reads just like every modern storm bust article saying uncertainty needs to be communicated better and it’s 17 years old!

Others are claiming that because a number of model runs showed sleet over the I95 corridor that the snow accumulation forecasts verified because it was still mostly sleet. This is cringeworthy too. What causes the difference between snow, sleet, freezing rain, and plain rain isn’t even meteorology 101, it’s junior high school weather. All of those precipitation types are related to one another in a winter storm and saying the snow forecast verified in this storm is like saying the snow forecast verified when plain rain fell. It’s simply not true.

If you have even a passing interest in weather, you should immediately understand this graphic. (from http://clintkywx.blogspot.com)

Many of the people claiming this are the same ones who are saying the euro verified. What they’re not saying is that the euro recently upgraded its model to include a wide range of precipitation types, even more so than any other model. Meteorologists and hobbyists were assuming the precipitation types factored into the snow maps from the euro being produced by different outlets and that if the snow maps were showing reduced snow it was because of sleet. These precipitation types weren’t factored into the snow maps, so when people thought the euro was forecasting eight inches of snow for our area when looking at the maps, these maps were actually a 10:1 QPF (Quantitative Precipitation Forecast) map of its output of snow and sleet combined. The euro’s later runs for the east coast verified pretty well in its QPF, but the snow forecasts from people did not as a result of this assumption. To give Dr. Ryan Maue of Weatherbell credit, he admits to this in this series of tweets here:

The one model that consistently forecasted the storm’s track and mixing along the I95 corridor was the NAM, but the NAM has performed so horribly the past five years or so that apparently nobody, even professional meteorologists, pay any attention to it at all anymore. This included me, as I didn’t factor the NAM into my forecast at all. Ironically, the NAM was built specifically to model the nuances of a storm such as this.

The NAM gets the win tag, this time. But, we still have our eye on you for shit forecasts, NAM.

To summarize the state of weather forecasting we’re in and some of the surrounding issues:

  • We have professional forecasters who rely too much on models.
  • The models and forecasts have stagnated, perhaps getting worse depending on the specifics, over the past decade.
  • The general public is as dumb as dogshit and is getting dumber every day, every hour.
  • Our politicians are dumb as dogshit and are getting dumber every day, every hour (probably because we’re governed by we, the people).
  • Actual news reporting and investigative reporting is dead and long gone in America.
  • It doesn’t matter if emergency managers, schools, etc. believe the forecast or not, they have to go with the worst-case forecast, otherwise their ass would be on the line.
  • Money has been poured into climate “science” for the past decade, rather than meteorology, and most meteorologists have said nothing.

It’s been my long-standing opinion that forecasting storms such as this won’t begin to improve until we exit the current climate science paradigm. So, here’s an idea that I wanna share: why don’t we admit that the doomsday predictions of catastrophic anthropogenic global warming and climate change haven’t come true and that it was all just politics, then start allocating some funding to improving the physics of models when it comes to weather, because short and medium range weather models are the primary tool used for weather forecasting. As someone who has paid good money in taxes over the past 20+ years with a keen interest in weather, knowing that money has been thrown into the black hole of climate science, I believe my voice should at least be heard.

But with the current mentality across America, I don’t see that happening and I see no reason why this broken record won’t repeat. Again. And again. The next one will occur either this coming hurricane season or next winter storm season, or during both. I’ve already got the image lined up for it.

We got about three inches of snow at my place before it changed over to sleet around 1am. It later changed over to a sleet and freezing rain mix. I’m not ultimately sure what type of ice accretions we got because I had to go to work early in the morning, but it was glazing badly as I left in the 6:00 hour. Seeing reports from around the area, this was a bust for most everyone in terms of snowfall and only out towards Fredneck did any real snow fall. My place seems to fall in line with those who are just northwest of I95 and those directly on it seemed to get 2.5 inches. The biggest bust in my forecast was northeastern MD, which seems to have only got about four inches. FAIL.

Totals:

Here’s the low as it fully formed:

As for the reason for the bust, the storm formed and tracked further west still from where both I and the models thought it would be and it tracked through eastern North Carolina and over the Norfolk area. In the morning it went just west of the mouth of the Chesapeake and on up over top of the Delmarva before exiting off the Delaware coast. Later in the afternoon it crossed Long Island. My understanding is that this was a bust for Philly to NYC to Boston too due to this track. I believe official forecasts were for snow to be measured in feet for those cities. FAIL.

Here’s what was the official forecast for New York City:

Here’s what it had changed to all of five hours later:


As for model performance, none of them got it right, but last night’s 0z NAM and Canadian forecasted the track correctly, but the storm had basically formed at that point so it shouldn’t have been that difficult. Not surprisingly, the 0z GFS still tracked the storm wrong and incredibly, so did the 06z GFS after the coastal had definitively fully formed. The fairly reliable (in the fist eight hours) HRRR indicated a colder solution in the middle of the night last night and that the mix would change back over to snow early in the morning and this was wrong too. The track on some model runs only a couple of days ago that took it well east were very wrong inside a time frame that is supposed to see improvement and further proof there’s still a long way to go to perfect the models. This after additional data from flights in the Atlantic were ingested into the models over the past couple of days to ensure forecast track and I’m sure additional weather balloon data too. FAIL.

The temperature did stay below freezing at my place for the duration of the precipitation heavy enough for accumulation, so I’ll give my forecast (and the models) that. It went below freezing at 9:05PM last night and went back above at 11:05am. Win.

Also to note: In one of the most perplexing things I’ve ever seen our local National Weather Service do, they dropped the Winter Storm Warning and put in a Winter Weather Advisory around 4:30am. I’ve never seen this in my ~32 years of watching the watches and warnings when there is ongoing heavy precipitation and the temperature is below freezing and not expected to go above, although they do it all the time when temps go above freezing. A winter storm warning is supposed to be 5 inches or more of snow or sleet within a 12 hour period, or enough ice accumulation to cause damage to trees or power lines, or a combination of both snow, ice and wind. We were certainly getting the combination of snow, ice and wind and were forecasted to through at least mid-day (the five inches of snow requirement has never counted when there’s mixing of icing). It was so odd, I had to go to the local NWS page to make sure they hadn’t changed the definition (they haven’t): http://www.weather.gov/lwx/WarningsDefined. I was completely baffled, but then I saw this article here and am thinking since they already had decided it was a fail before the storm started, they just went ahead and downgraded it regardless.

Later during the morning, tree limbs being down were reported everywhere. Here’s one in Canton that is on my drive to work:

Also to note: Only two weeks ago it was loudly declared by the CWG spring had arrived way early and that it could/would be an early record breaking cherry blossom year. A conveniently non-paywalled Washington Post article highlights this:

A couple of days later the NPS officially said the festival will kickoff tomorrow:

Now they’re saying they could be 90% dead:

LOL. Oops. FAIL.

The biggest threat of the season is almost here and already virtually all of the weather outlets have downsized their predictions pointing to a bust scenario being more likely. Currently the National Weather Service believes 5-7 inches will fall in Baltimore and 4-6 inches in DC:

This is because significant mixing with sleet is being portrayed by the models due to their belief the storm is going to directly hug the coast now as it passes by us. Here’s a loop from the High Resolution Rapid Refresh (HRRR) model showing an extended period of sleet for I95 in the Mid-Atlantic. This is one such possibility, but usually the sleet band is slightly thinner than what is being portrayed:

But there’s still room for both boom and bust – the first flurries started flying around 6pm in the Baltimore area, a solid three hours ahead of model forecasts, so already the storm is behaving a little different than expected. It may mean nothing, but no storm has ever gone exactly as a model predicted even at event time. And interestingly, the “expect at least this much” (10 percentile) and “potential for this much” (90 percentile) graphics have changed to both lower and raise totals across the area.

The 10 percentile graphic shows nothing for the I95 corridor now:

But the 90 percentile graphic shows an even larger area of 18+ inches intersecting with the Baltimore City line:

So while the National Weather Service has lowered expectations overall, there’s actually more uncertainty in their forecast than there was yesterday.

As for the models, the 12z euro gave 7 inches down the I95 line:

The 18z GFS gives 4-6 inches:

The 12z Canadian gives big amounts, but much of this is sleet being translated as snow on the map:

As for my forecast if I could change it, I would downsize too with a few inches of snow followed by significant sleet. Wednesday is going to be very cold behind this system with temperatures not getting above freezing, a rarity for mid-March. That’s all I’ve got.

I issue a final forecast 24 hours prior to the beginning of an anticipated snow storm. Beyond that is just nowcasting IMO.

  • Directly along I95 from DC to Baltimore: 4-8 inches with the lighter amounts the further south/east one lives
  • North and west of I95 from Owings Mills to Rockville to Centreville: 8-12 inches
  • Northeastern MD (Harford and Cecil counties): 9-13 inches
  • Southern Anne Arundel county, southern and eastern PG county, north, central and western Charles County: 2-4 inches
  • Southeastern Charles, Calvert, and St. Mary’s counties: 0-2 inches

There’s no real changes in amounts or my thought process from my previous forecast. It’s a certainty that schools will be closed everywhere Tuesday except southern MD given the hype. As for OPM, I’m not sure what their thinking will be. Timing of first snow flakes will be sometime between just after rush hour Monday and prior to 11pm. Heaviest precipitation rates will be between midnight and 8am Tuesday and in those spots where it is snowing, rates could be two inches per hour or slightly higher for several hours.

A storm like this makes me happy I’m not a professional meteorologist. The issues noted in my previous post could lead to a very sharp gradient of snow totals and it’s within the realm of possibility that gradient could be quite large. It’s possible that someone somewhere along the I95 corridor ends up with a few inches of slush while someone 5 miles to the north and west as the crow flies receives 18 inches, although I don’t think the difference will ultimately be that dramatic over such a short distance. The 12z GFS showed no mixing of precipitation for the I95 corridor and the 18z was very similar:

But I find it unlikely to be entirely snow for everyone except southern MD like it shows. Our local National Weather Service agrees that mixing is a distinct possibility over the I95 corridor:

The 12z Canadian showed sleet for a time along the I95 corridor and I find that to be the most likely scenario and sleet would cut down on the accumulation totals:

The 12z euro gave 8-10 inches over the I95 corridor, because it was the strongest bomb of them all with its coastal front overtaking the I95 corridor as well:

Additional Notes:

  • Even if the moderate snow totals in my forecast were to verify, it still wouldn’t bring this winter up to normal snow fall. The map below shows the paltry snow for our area this winter down at the bottom.

  • As with any storm undergoing bombogenesis, thundersnow is possible. I wouldn’t be surprised if there are a couple of reports on social media of thunder being heard if the bombing occurs like the models say it will.
  • If an area of convective snow breaks out in our area, I wouldn’t be surprised if the National Weather Service upgraded to a blizzard warning. Right now most of the models keep the wind gusts just under blizzard criteria, but convection would bring stronger winds down to the surface. This scenario (where the NWS was originally not going to issue blizzard warnings, but then had to) happened in the third blizzard of the 2009-2010 season. The 12z euro’s wind gusts would meet blizzard criteria for portions of the area if snow were falling.
  • I finally bought a semi-professional weather station a few weeks ago when AcuRite had a good deal going. I figured today was as good as any day to set it up with the possible snow storm coming. It’s supposed to be able to connect to weather underground but the wunderground site is giving me no response when trying to create a site. I’ll try again tomorrow, but I’m not concerned whether it’s on wunderground or not.

For the record, the local NWS “most likely” scenario at the time of this writing (6:00pm, 03/12) is:

And the “expect at least this much” and “potential for this much” are:

 

The mid-day model runs diverged fairly significantly with the snow possibility for Monday night into Tuesday. First up was the GFS, which had a coastal hugger and inland track once it got up to northern New England. The hugging of the coast resulted in the rain/mix line going over top of the I95 corridor for the Mid-Atlantic in the middle/end of the storm resulting in what would likely be a 6 or 8 inch slop when it was all said and done. The Canadian came in next and its 12z run took the storm east with virtually no snow for the I95 corridor in the Mid-Atlantic. The euro then came in splitting the difference and giving all snow to the I95 corridor and southern MD with accumulations likely over a foot.

Out of the three, I find the GFS and euro scenarios to be the most likely with the Canadian being the more unlikely of the three. In fact, the Canadian’s upper level energy was weaker than the other two and its model run formed a rather odd looking double barrel of low pressure oriented north-south out over the Atlantic. It was odd given the scenario and the surface low pressures splitting like this and weaker upper level energy may have been a result of the model initializing poorly. The 18z GFS was nearly identical to the 12z run, but it made a small step towards the euro’s solution in the upper levels of its model run.

Perhaps the best sign for snow lovers is the euro ensembles that came in the middle of the afternoon. The ensembles gave Baltimore a 50% chance of a foot of snow or more. This is a quite high probability this far out and would be a huge fail for the euro model suite if no snow were to fall for Baltimore.

My decision on whether I think snow is a probability or not will come Saturday afternoon.

The operational euro kicked things off on Saturday, March 4 showing two feet of snow for the March 12-13 time frame getting those who don’t know better excited. The well-known-to-be-outrageous euro control model of the ensembles even produced 40 inches. Then a number of professional and amateur meteorologists got actual raging boners for the Sunday, March 12 Mid-Atlantic snowstorm possibility this past Monday, March 5 based on near-consensus among the GFS, Canadian, and euro giving 6-10 inches in their model runs.

Then the possibility completely faded within 24 hours worth of model runs and by Tuesday mid-day, it was obvious that the upper level and surface patterns would send the storm south and the meteorologists all went completely flaccid.

The professional meteorologists are lucky the models didn’t continue to advertise the storm. All of the pent-up anxiety over a no-snow winter with many busted winter forecasts could have resulted in a March, 2013 snowquester-type bust where meteorologists went with the models’ wishcasting despite their obvious shortcomings and borderline-cold March airmass.

What this winter has proven is that absent an absolute textbook upper level pattern setup, like with the blizzard of January, 2016, the models (euro included) still suck balls beyond five days and often beyond only three days. Given that the models have sucked beyond 3-5 days for over 20 years now, I blame funding having gone to the great global warming (or climate change) fraud, rather than weather prediction, for the stagnation. That and the fact that I believe weather and climate is very complex, something that the prevailing weather and climate science paradigm insists is very simple. Until the paradigm shift finally occurs, I expect more of the same.