28 June Extreme Flooding in Herkimer County
(click on any image to enlarge)
An active stretch of weather was anticipated a few days in advance as forecast discussions as early as June 25th and 26th mentioned the possibility of a strong wave of low pressure impacting portions of NY and New England with very heavy rainfall and more flooding. The HWO was specifically updated at 1243 am WED June 26th to mention the possibility of a Flood or Flash Flood Watch in the Days 2 to 7 segment. A Flood Watch was issued later that afternoon with high confidence for main stem river and flash flooding from an anomalous low pressure system Thursday late afternoon into Friday morning. The northern and western portion of the forecast area looked most vulnerable for flash flooding due the wet and saturated soil conditions from the preceding 5 days, especially the western and central Mohawk Valley. Southern Herkimer was hit especially hard on Sunday, June 23rd, and Tuesday, June 25th.
A strong short-wave trough approached the forecast area from the lower Great Lakes Region and Ohio Valley Thursday morning. An anomalously deep cyclone approached the region Thursday night from the Ohio Valley and Pennsylvania with very heavy rainfall on the north and east side of the 995 hPa low. This low pressure system was 2 to 4 standard deviations /STD DEVs/ deeper (lower) than normal. Precipitable water values were forecasted to be 2 to 3 STD DEVS above normal. The 850 hPa southerly /+V component/ wind anomalies were impressive in a 12-hr window being 3 to 5 STD DEVs above normal tapping Gulf and Atlantic moisture. The 12Z GEFS on 26 June also had 850 hPa moisture flux values in excess of 5 STD DEVs above normal. The easterly /-U component/ wind anomalies were also impressive for a short period of time at 3 to 5 STD DEVs above normal over west-central NY. The trends would continue with the 18Z and 00Z GEFS. Localized rainfall enhancement off the terrain was expected and warm cloud collision coalescence processes /high warm cloud layer depths/ would allow for excellent precipitation efficiency in the moisture rich atmosphere. The radar data will have to be checked for low echo centroids with the maximum reflectivities below the freezing level. The heaviest rainfall was expected between 22Z THU and 16Z FRI. The strongest QG lift or upward vertical motion would come ahead of the warm front and with the surface cyclone. There were some indications of a coupled/dual jet structure with the poleward jet streak departing over southern Quebec and the St. Lawrence River Valley and an approaching equatorward jet streak from the Appalachians and Ohio Valley. The left front quadrant of the jet streak would help deepen and intensify the surface wave Friday morning.
There was some question whether a dry slot would blast through the region cutting down on the QPF across the hydro service area. 1 to 3 inches of rainfall was expected in the 12 to 18-hr time frame with some localized 3 to 5 inch amounts. Down sloping was also a concern with the strong easterlies. Embedded convection was also possible with negative Showalter values indicated by several sources of the guidance. WPC a few days in advance had a "Slight Risk" for exceeding flash flood guidance in a 1/3/6/12-hr time frame.
Prior to nightfall some isolated or scattered convection popped up around the area with some minor flooding reported in eastern Greene County. Some heavier cells also formed in Warren and Columbia counties...and flood advisories were used to address this convection. The convection in Warren County may have saturated the soils enough for flash flooding that was hard to detect Friday morning in Bakers Mills/Johnsburg. The morning rainfall was widespread but only amounted to one to two inches at most. Flood advisories were used overnight in this area, and signals were not in FFMP for flash flooding in north central Warren County.
The main attention was on very heavy rainfall Thursday night and Friday morning over the western periphery of the ALY HSA, most specifically the west-central Mohawk Valley and the western Adirondacks. Portions of the BGM HSA were getting hit extremely hard with heavy rainfall, especially the western Catskills and Pike County in PA. This rainfall quickly moved through the eastern Catskills with a few inches occurring, but no flooding. The Mohawk Valley was a different story, and the forecast staff issued a flash flood warning with multiple inches of rainfall expected in southern Herkimer, Fulton, and Montgomery counties shortly after midnight.
The Flash Flood warning issued for southern Herkimer, western Fulton, and western Montgomery counties was issued based on the wet antecedent conditions, and the very heavy rainfall moving in from the south and east. The warning was issued for 6 hours initially. The polygon was made large enough to intentionally include Canajoharie in Montgomery County. Some of the locations had a quarter of an inch to an inch of rainfall the day before. A FFS was issued at 431 am, as spotter in Columbia Center reported over 3 inches of rainfall so far, and there were growing concerns of significant flash flooding based on a 911 Emergency Management call with numerous roads washed out and impassible with homes threatened. Details had to wait, since the 911 center was busy and chaotic was the significant flash flooding. The staff talked about a flash flood emergency, as we tried one of our spotters, but could not get a return call. Around 5 am emergency management talked with the lead forecaster. Conditions were perilous, and the staff issued a flash flood emergency for several communities. It was decided to use the same polygon for the flash flood emergency and to extend until 1230 pm. The flash flood emergency was typed into the warning. Local fire officials let the staff know that swift water rescues were underway in the chaotic situation. This was included in the emergency statement/warning. Several main river warnings were issued. The Kast Bridge one was a little late, but complicated based on the anomalous nature the data came in. See below...
What was learned from this event:
In Warren County, the amount of rainfall was less overnight, and only an FLS was issued there. Not sure why the flash flooding was concentrated in Johnsburg. Flash flooding was extremely isolated, water surrounded a home and a family was evacuated. Still not clear on cause, since radar rainfall estimates don't seem to support this.
The heaviest bands of rain that occurred to the north of the cyclone may have had characteristics of mesoscale banding similar to what is seen in cold season rain/snow events. Frontogenesis fields will have to will have to be investigated to confirm this hypothesis. Also, portions of central/southern NY including Herkimer County were in the area where the heavy rain bands set up and pivoted late Thursday night into Friday morning that resulted in the devastating flash flooding.
Above: Slight risk of flooding issued by the Weather Prediction Center.
Above: Graphic of 7 day departure from normal precipitation showing the very wet antecedent conditions in Herkimer County.
Above: Soundings from Albany. NY (KALB) at 0000 UTC 28 June (left) and 1200 UTC 28 June (right).
Above: Radar mosaic loop (courtesy of College of DuPage). Note the vorticity center that tracked through central NY.
Above: Observed stage 4 precipitation analysis showing the extreme rainfall in central NY including Herkimer County.
Above: River levels at Little Falls and Kast Bridge showing the extreme runoff into the rivers.
Above: Pictures around the Erie Canal and canal street.
Above: Pictures from around Fort Plain.
Above: Pictures from around Herkimer.
Above: Pictures from around Ilion.
Above: Pictures from Little Falls.
Above: Pictures from around Mindenville.
Above: Pictures from around Mohawk.
Above: Road washouts in the area.
Above: Pictures along the NY State Thruway and West Canada Creek.
Above: Pictures from a separate flooding incident around Johnsburg, near Warrensburg.