1. REMOTE SENSING: DATA COMBINATION AS A KEY FOR STORM NOWCASTING
DEPARTMENT OF ATMOSPHERIC PHYSICS
CHARLES UNIVERSITY
REMOTE SENSING: DATA COMBINATION AS A KEY FOR STORM NOWCASTING
Michaela Valachová, Hana Kyznarová, Petr Novák, Martin Setvák
9th European Conference on Severe Storms
Pula, Croatia, 20 September 2017
Central Forecasting Office, Prague

2. Suomi-NPP/VIIRS sandwich 2013-06-20 11:05 UTC
outline
motivation of this work
data used in this study (source & software)
lightning detection network (CELDN & R)
satellites (EUMETSAT & McIDAS-V, Python)
radars (CHMI & CELLTRACK, R)
ESWD (ESSL & R)
example: severe vs. non-severe storm
summary, acknowledgements
Suomi-NPP/VIIRS sandwich :05 UTC

3. forecaster’s point of view
convective storms are challenging
where and when will storm evolve ?
how dangerous will it be ?
how long will it last ?
remote sensing is available
information every 5 min
years of experience and data
independent sources
Forecaster has the confidence to issue warning in time !!!
FORECASTERS IN THE ROOM ???
storm weakening or strengthening ???
We already have what we need, no new data… Our research aim to fulfill remote sensing potential and meet the forecasters’ requirements for a better nowcasting tool
wide variety of nowcasting systems and models across Europe → overwhelming, non-transparent, inconsistent

4. possible utilization
electrification, dynamics and microphysics connected → changes visible in all remote sensing data → NOWCASTING
Electrification process, Cloud Dynamics, Microphysics in the cloud… USEFUL INDEPENDENT SOURCES OF INFORMATION ABOUT STORM SEVERITY
clear, when the storm strengthen or weaken
highlight distinct changes in cloud dynamics, understanding microphysical processes in Cb

5. remote sensing
different data sets in Europe

6. RADARS microphysical properties and dynamics
Polarimetric Doppler radars (upgrade in 2015)
C band (λ ~ 5 cm), 12 elevations
resolution 1×1 km (whole domain)
many useful applications:
CELLTRACK, COTRACK
CELDN strokes
PrecipView, WarnView
Circles depict the maximum measurement range of individual radars. Irregular lines depict maximum range, where the height of the radar beams 1.5 km and less above ground.

7. Operational output of CELLTRACK (JSMeteoView)
RADARS - celltrack
reflectivity cores tracking algorithm
developed in CHMI (Hana Kyznarová)
For the presented study:
no tracking, just identification of cores
characteristics of cells:
threshold of 44 → 20 dBZ (isolated storms)
parameters:
AREA, VOL, VOL44
HP, POSH, MESH, VIL
identification of reflectivity cores: a single threshold of 44 dBZ applied to a maximum reflectivity field for identification of reflectivity cores as an approximation of convective storm cells
avoid the situations similar to this figure 
visualisation software JSMeteoView
Operational output of CELLTRACK (JSMeteoView)
Kyznarová H., Novák P. (2009): CELLTRACK – Convective cell tracking algorithm and its use for deriving lifecycle characteristics, Atmospheric Research, vol. 93

8. lightning detection
CELDN (Central European Lightning Detection Network)
part of EUCLID, operated by Siemens AG
operatively used in CHMI
more or less as “ground truth” or addition to the radar data – not used itself as another source of information
“Total lightning is the best early indicator of a strengthening updraft within a storm.” on European Conference on Severe Storms 2013, Helsinki (Schultz, Petersen, Carey – in Weather and Forecasting)

9. lightning detection microphysical properties, strength of updraft
every stroke: type (CC, CG), time [ms], location, current amplitude estimation [kA] and polarity
detection efficiency: about 90 % or higher for CG
location accuracy: about 1 km for CG
uncertain estimate of current amplitude ~ tens of %
no stroke clustering into flashes
Information about clustering strokes into lightning are not provided within CELDN and no such algorithm is used in CHMI. Therefore strokes are not grouped and are treated individually as detected. However, for a detailed analysis of individual convective storms only relative comparison of the electrical characteristics throughout the life cycle and physical nature of this phenomenon are important.

10. Aqua/MODIS 2013-06-20 12:25 UTC; hail occurrence at 12:26 UTC
satellites
microphysical properties and dynamics
geostationary:
Meteosat/SEVIRI RSS
polar orbiting:
Suomi-NPP/VIIRS
Aqua/MODIS
CloudSat/radar and CALIPSO/lidar, imagers
infer information about storm intensity, possible severity or internal structure
Aqua/MODIS :25 UTC; hail occurrence at 12:26 UTC

11. severe weather reports
reports from ESWD operated by ESSL
quality control: QC0+, QC1, QC2
time uncertainty up to 15 min
only “positive events”
NO GROUND TRUTH

12. Severe OR NON-SEVERE ?

39. geometric center of the storm cell detected by CELLTRACK

48. summary - satellites indicators showing storm development
rapid anvil spread
rapid cooling of the cloud-top
features on the cloud-top
distinctive overshooting top
ice plume
small ice particles
cold-U or cold-ring shape in IR-BT

49. summary - radars lifetime and track of radar cell
radar derived parameters:
radar reflectivity, height of the lowest reflectivity
HP, POSH, MESH, VIL

50. summary - lightning pulsation in number of strokes
multimodal histograms
two or more processes, spacing ~ min
the first OT ~ the first peak in all strokes
severe weather occurrence
abrupt increase of CC/TL as a precursor
amplitude of strokes
non-severe storms mostly low amplitudes (< 15 kA)
the first CG+ amplitude > 20 kA when a significant change inside the storm
CC increase, plume formation, OT more frequent, radar reflectivity increase

51. “Hessen storm”

52. acknowledgement
Support, motivation, inspiration: Patrik Benáček (CHMI) Hana Kyznarová (CHMI) Katrin Wapler (DWD) André Simon (OMSZ) Justin Sieglaff (CIMSS UW) John Cintineo (CIMSS UW) David Rýva (CHMI) Data source: CHMI, EUMETSAT, Siemens AG ESSL and all active spotters

54. references
Wapler K. (2017): The life-cycle of hailstorms: Lightning, radar reflectivity and rotation characteristics, Atmospheric Research, vol. 193 Wapler K., Harnisch F., Pardowitz T. et al. (2015): Characterisation and predictability of a strong and a weak forcing severe convective event – a multi-data approach, Meteorologische Zeitschrift, vol. 24 Cintineo J. L., Pavolonis M., Sieglaff J. et al. (2014): An Empirical Model for Assessing the Severe Weather Potential of Developing Convection, Weather and Forecasting Sieglaff J. M., Hartug D. C., Feltz W. F. et al. (2013): A satellite-based convective cloud object tracking and multipurpose data fusion tool with application to developing convection, Journal of Atmospheric and Oceanic Technology, vol. 30 Bedka K., Wang C., Rogers R. et al. (2015): Examining Deep Convective Cloud Evolution Using Total Lightning, WSR-88D, and GOES-14 Super Rapid Scan Datasets. Weather and Forecasting, vol. 30 Dworak R., Bedka K., Brunner J., Feltz W. (2012): Comparison between GOES-12 Overshooting-Top Detections, WSR-88D Radar Reflectivity, and Severe Storm Reports. Weather and Forecasting, vol. 27 Schultz Ch., Petersen W., Carey L. (2011): Lightning and Severe Weather: Comparison between Total and Cloud-to-Ground Lightning Trends. Weather and Foresting, vol. 26 Novák P., Kyznarová H. (2011): Climatology of lightning in the Czech Republic. Atmospheric Research, vol. 100 Kyznarová H., Novák P. (2009): CELLTRACK - Convective cell tracking algorithm and its use for deriving lifecycle characteristics. Atmospheric Research, vol. 93 Novák P. (2007): The Czech Hydrometeorological Institute's severe storm nowcasting system. Atmos. Research, vol. 83