Disadvantages of False Color Composite Differences
Total Page:16
File Type:pdf, Size:1020Kb
Gina Loss and David Bernhardt ‐ NOAA/NWS, Great Falls, MT, Kevin Fuell ‐ UAH/NASA SPoRT, Geoffrey Stano ‐ ENSCO, Inc./NASA SPoRT Introduction Findings River Ice Monitoring Samples The rural nature and low population density of Montana Advantages of false color composite create difficulties in the evaluation of snowfall and river ice. Assessment Samples 2004 Case Assessment Remote sensing provides the best opportunity to appropriately • Complete overview of snow cover extent Snap‐shot comparison of snow and cloud cover as seen by River Ice determine the transition of snow cover and river ice in remote February 19, 2004 − High spatial and spectral resolution MODIS visible, natural color and false color. Valley Fog areas, however, visible imagery does not clearly differentiate Widespread snow cover over − Information for areas void of surface data P o p Mi la lk r northeast/east Montana. Ri − Indicates primary areas of concern when complemented snow from clouds. ver R i Snow in v Snow in Snow e Bare r Observations show areas with with supplemental data from ground measurements The Great Falls WFO (Weather Forecast Office) has been Forests Forests Ground ver i Ri snow depths of 2 feet or more sour − Often eliminates need for field surveys during potentially assessing the use of the MODIS (Moderate Resolution Imaging Mis Clouds River and Rotting River and snow water equivalents of 6 dangerous situations Spectroradiometer) false color composite created by the Short‐ Lake Ice and Lake Ice to 8 inches. Significant flooding term Prediction Research and Transition (SPoRT) Center at Snow • Observation of snow cover transition Clouds of Milk and Poplar Rivers Bare − Speed and extent of snowmelt provide insight to areas NASA/Marshall Space Flight Center to monitor snow cover and Ground Bow‐tie forecasted with melt. Extensive River Ice Rotting river ice. It is expected the false color composite can lead to Snow in Effect Lake Ice with possible flooding concerns briefing of emergency Forest Lake Ice improved assessments of flooding potential during post event management officials begins. • View of ice on rivers and lakes conditions where rapid melting and runoff are anticipated. Snow − Ice formation and degradation provide insight on locations of possible ice jams and related flooding Snow March 4, 2004 Cirrus Snow in Snow in Clouds Snow melt beginning. Imagery • Corroborates information provided by NOHRSC (National Forests P Clouds Forests o p Mil la MODIS False Color Composite k R r i Operational Hydrologic Remote Sensing Center) snow models ver R shows significant loss of snow i v e r MODIS false color imagery composites one visible and two Montana snow and river ice observations and comparisons using MODIS false color imagery February 21, 2008 (left) and April 12, 2008 (right). south of Missouri River. No iver infrared channels to highlight features with infrared signature uri R isso flooding reported, however Disadvantages of false color composite differences. Snow and clouds have reflective differences above M snow water content was not as • Cloud cover blocks view of surface 1.4μm, especially near the 1.6μm and 2.13μm MODIS channels. Visible image — White snow and clouds blend as they would if great in this area as in northeast − Visible and infrared spectrums Compositing a visible channel with these two infrared channels viewed by the human eye. Assessment of Flooding Potential Montana. Concerns for flooding − Significant snowfall/snowmelt can occur under cloud creates an image that clearly distinguishes snow and clouds. continue for northeast cover and full extent may not be viewable for days Snow event Montana. Co‐op Observations • Images per day limited − Polar‐orbiting satellite provides only a few images daily Monitor change Clouds in stream flow March 11, 2004 • ‘Bow‐tie’ effect blurs images P Melt not occurring uniformly. − Satellite field of view overlap produces data repetition at Snow o Examine snow cover p Mi la lk R r Bare iv er R image edge i Snow south of Missouri and USGS Streamflow v e Ground r − Algorithm to remove this effect will be available soon Milk Rivers virtually gone Snow in ver i Ri sour Forest Mis while widespread snow remains north. Skewed spatial NWS Streamflow and Forecast Conclusions SNOTEL Reports distribution of melt and influx Snow Illustrates stream Faster and more Satellite Imagery of water to rivers expected to The Great Falls WFO did find the MODIS false color image Cirrus useful in monitoring snow cover and river ice extent. It was used Clouds flow/snowmelt convenient reduce chances of significant river flooding. as a complement to other tools such as the NOHRSC snow water relationship equivalent data and soil saturation and temperature conditions. It Typical reflectance for naturally occurring features. Spectrally, Satellite Imagery NOHRSC Model Snow Water was felt the imagery did lead to improved assessments of snow is different from clouds at wavelengths greater than 1.4μm Equivalent March 18, 2004 flooding potential during post snow and ice events where rapid and particularly near the 1.6μm and 2.3μm MODIS channels. Natural color image — Composite of three visible channels. Imagery and observations show melting and runoff were anticipated; the imagery provided P o White snow and clouds blend, again, as they would if viewed by the p Mi la lk r Facilitates Ri information for rural and remote areas, allowing staff to quickly Benefits Examine NOHRSC ver R melt has slowed. Snow now i v human eye. e r monitoring 2‐D provided by products gradually retreating north of determine and focus on those areas of greatest concern. Further, Comparison to Other Satellites r ve i Ri sour the Missouri and Milk Rivers. the imagery often eliminated the need for field surveys not only MODIS offers an increase in spatial as well as spectral change of snow MODIS Corroborates NOHRSC Model Snow Depth Mis Monitor change in Slowing of melt will allow more saving time and resources but also keeping personnel safe during resolution over GOES (Geostationary Operational Environmental cover False Color NOHRSC snow snow cover and gradual influx of water to potentially dangerous situations. Satellites) and AVHRR (Advanced Very High Resolution Imagery model products Clouds rivers, further reducing the Radiometer). The result is a detailed view of surface and snow water Snow chance of significant flooding. atmospheric features. equivalent Bare References Ground Adolphson, J., et. al., 2005: “Minor Flooding on the Milk River After an Extreme Great Falls is at a relatively high latitude and at the edges of Winter in Northeast Montana” Preprints, 19th Conference on Hydrology, Amer. Snow in Saves human both GOES East and West. The remapping of data results in a Forest NOHRSC Modeled Snow Melt Meteor. Soc., San Diego CA, 4.3 north‐south and east‐west blurring of details. GOES imagery can resources and Hall, D. K. and G. A. Riggs, 2007: “Accuracy assessment of the MODIS snow Provides easily March 25, 2004 products”, Hydrol. Process., 21, 1534‐1547 be looped to show stationary snow features vs. moving clouds, increases safety NOHRSC Aerial Survey P Snow cover north of the o Jedlovec, G. and P. J. Meyer, 2005: “MODIS Composite Imagery for Snow Detection p understood Mi la however, there are still issues with semi‐stationary clouds such as lk R r Snow iv er R Missouri and Milk Rivers ‐ Training Module”, http://weather.msfc.nasa.gov/sport/sport_training.html NOHRSC Flight Lines i v Cirrus e r fog, lenticular formations, and orographically induced stratus. Clouds briefing tool nearly gone. During the course Miller, S., 2003: “High/Low Cloud and Snow Discriminator – Focus Tutorial” NRL iver Monterey, Marine Meteorology Division, http://nrlmry.navy.mil/sat_training/ uri R isso of the snowmelt event, only MODIS AV H R R GOES M high_low_cloud/modis.html Inform emergency minor flooding occurred. Spatial Resolution (meters) 500 1000 1000 management Channels used to produce snow products 7 3 1 broadband False color image — Composite of one visible and two infrared Request survey Images per day 1‐2 3‐6 20 channels. Snow (red), clouds (white) and bare ground (blue‐green) of remote areas Comparison of features with MODIS, AVHRR and GOES are distinct. .