NOAA Satellite Snow Cover Extent Climate Data Record at the Half Century Mark David A. Robinson Rutgers, The State University of New Jersey Piscataway, NJ WMO Global Cryosphere Watch: Snow Watch Team – 2nd Session Columbus, Ohio | June 13, 2016 Research supported by Global Science and Technology, Inc. at NOAA NCEI

NOAA Visible Weekly SCE Climate Data Record Weekly charts Digitization Daily IMS Nov 1966 Oct 1972 May 1975 1980-81 1990s May 1999 ESSA, NOAA, GOES Series Weekly 190 km Digitized METEOSAT & GMS added Reanalysis of 1966-71 1988-89 Feb 1997 2004 IMS 24 km IMS 4 km Interactive Multisensor Snow & Ice Mapping System Dec 2014 IMS 1 km

NW Arizona

Quebec/Labrador

10 March 2015 Mauna Kea, Hawaii 22 February 2015

NOAA Visible Weekly SCE Climate Data Record Specifications Binary (snow / no snow) over NH land surface 88 × 88 Cartesian grid on polar stereographic projection 190.6 km resolution at 60°N Weekly temporal resolution October 4, 1966–present Inputs to CDR October 1966–May 1999: primarily visible satellite imagery from multiple instruments After May 1999: Interactive Multisensor Snow and Ice Mapping System (IMS) SCE derived from multiple sources by trained analysts NH SCE CDR simplified processing flow diagram

Interactive Multisensor Snow and Ice Mapping System (IMS) 1 km, 4 km & 24 km Northern Hemisphere Analysis Snow & Ice Cover Produced daily at U.S. National Ice Center Pre-Processing Indirect Sources Analyst Derived Output Satellites GOES (E & W) MeteoSat (MSG & 7) MTSAT NOAA Automated Snow & Ice AVHRR (Channels 1 & 3) MODIS (Channel 8) ASCAT AMSU (Derived snow, ice, rain) Other Sources Radar Models Surface Observations Webcams Buoys Charts

Northern Hemisphere Continental Snow Cover 10 January 2012 Extent Departure (blue: positive; red: negative)

Northern Hemisphere Continental Snow Cover January 2012 Extent Departure (blue: positive; tan: negative)

Swings from most to least extensive SCEs occurring within months Departures Feb 2010 May 2010 3rd most extensive 1st least extensive

NH Snow Cover Extent: 1966-2016 Departures derived from 1981-2010 mean, plotted on 7th month

NH Spring (MAM) SCE Departures 2016 is lowest on record at 27.14 mil. sq. km (-2.65 anomaly), 1981-2010 mean is 29.79 2nd and 3rd lowest are 1968 (27.34) and 1990 (27.43) respectively Departures derived from 1981-2010 mean

May NH Snow Extent Anomalies: 1967-2016 4th lowest May on record May 2016 Departure from 1981-2010 Mean

NOAA SCE compared with station observations fractional coverage of snow North America: Compare one degree cells from 53°W – 168°W longitude and 20°N – 71°N, not including any coastline and with at least 5 stations. Extent adjustment to “reach” IMS era: +5.23% 1965-80, +2.75% 1981-98 (NOAA maps underreporting in the pre-IMS era)

Time to start generating an IMS CDR: 17 years of operational IMS output and growing Operational since November 1998 24 km resolution at 60°N Daily temporal resolution Bring IMS SCE output up to CDR standards

Mean IMS SCE January Percent of days snow covered Total 526 days, lowest value 1 day (0.19%)

Mean IMS SCE April Percent of days snow covered Total 508 days, lowest value 1 day (0.197%)

Mean IMS SCE May Percent of days snow covered Total 526 days, lowest value 1 day (0.19%)

Mean IMS SCE Percent of days snow covered April May

Mean IMS SCE Percent of days snow covered January April

Mean IMS SCE Percent of days snow covered January May

2006 2008 2010 2012 Columbia Basin SCE April 1 Source: 4 km IMS daily

Columbia Basin: IMS SCE vs. Discharge April-July streamflow volume in the Yakima basin as a function of percent snow cover for each month [right], including only the 5 key 24 km grid points in the basin [above]. We determined that by looking at certain regions within a subbasin that experience more variability in snow cover than others, that a SCE – discharge signal emerged. These were mainly grid points at moderate elevations where greater variability in snow extent is seen within a season and more notably, inter-annually. Five such 24 km cells were identified (out of 33) in the Yakima basin. By looking at certain regions within a subbasin that experience more variability in snow cover than others, a SCE – discharge signal emerged. These were mainly grid points at moderate elevations where greater variability in SCE is seen within a season and more notably, inter-annually.

Columbia Basin: IMS SCE vs. SNOTEL SWE Green Lake SNOTEL station within the Yakima basin. SWE (blue line) (inches) at the SNOTEL station and the days when the 4 km cell was mapped as snow covered in the IMS product (grey vertical bars) from 2004-2015. Within the Yakima subbasin a preliminary study examined whether there was a relationship between SCE depicted within IMS 4 km grid cells and the snow water equivalent (SWE) measured at SNOTEL stations within the cells. The graph [above right] shows a strong relationship between SWE and IMS melt out. In other cases, at more heavily forested sites the IMS depicted melt out prior to the loss of SNOTEL SWE. The graph shows a strong relationship between SWE and IMS melt out. In other cases, at more heavily forested sites the IMS depicted melt out prior to the loss of SNOTEL SWE.

Issues warranting consideration Local, regional, hemispheric: Timing of season Length of season Average and extreme depth Average and extreme snow water equivalent Associated impacts physical societal

Pursuing answers Ongoing/improved monitoring in situ satellite integrated Change detection/attribution investigations Ongoing assessments of snow cover within the system data and model driven Cooperation with social scientists

Thanks Dave Robinson david.robinson@rutgers.edu snowcover.org Somerset, NJ, 21 March 2015 NASA/GSFC/Suomi NPP White Marble – 26 May 2012