The Utility of Cloud-to-Ground Lightning for Debris Flow Warning in the Belluno Province of Northeastern Italy

ABSTRACT: Lightning detection networks now flourish around the globe and with the expansion of these networks regions prone to particular hydrometeorological hazards can be targeted for analysis with the goal of providing better hazard recognition and more prompt hazard warnings. In this context the Italian lightning network Sistema Italiano Rilevanmento Fulmini (SIRF) has allowed for analysis of debris flow initiation in the Belluno province of the Veneto Region in northeastern Italy. The study presented here uses cloud-to-ground (CG) lightning flash information as an indicator of convective activity in the province and as a warning parameter for localities (the study site) within the province. The study site for this research is the Acquabona watershed in the Dolomites of northeastern Italy. The watershed, which includes the town of Cortina d'Ampezzo, experiences one of the highest frequencies of debris flow activity on the European continent.

The Acquabona watershed has been the focus of debris flow research for more than a decade and is outfitted with research infrastructure that includes rainfall gages, channel bed pressure transducers, and video cameras. Debris flow initiation data consisting of rainfall intensity, time of excess runoff, time of debris entrainment, and debris volume in channel are logged and archived by researchers at the University of Bologna (Italy). A portion of this data set is published in Gregoretti and Fontana (2008). The Acquabona archive was used in this study to identify rainfall events that produced active debris flows in the watershed. Identifying six such events the researchers conducting this study acquired addition data sets including CG lightning flash data (SIRF) and satellite-derived cloud-top temperature (CTT) data (EUMETSAT).

Using methods developed by Holle and Bennet (1997) and Underwood and Schultz (2004) the CG lightning data were partitioned into five minute intervals and the following parameters calculated: total CG flashes, peak CG flash rate; continuous CG flash intervals; percent positive flashes; peak current of CG flashes; and average current of flashes. The CCT data were analyzed along three north-south and three east-west transects proximal to the debris flow study site. The 30-minute CTT images provided information including: minimum transect temperature, transect temperature range, and mean transect temperature.

Utilizing the parameters listed above over the six events selected the study was successful in defining the most “valuable” spatial scale for using CG flashes to identify intense local scale rainfall in at the study site (17662km2). The study also found that in using this spatial domain the mean peak five-minute CG flash rate (for the six events) preceded peak rainfall at Acquabona by an average of 21 minutes. Minimum CTT in the nearest transect to the study site was 230K. No null cases were employed in this study so there are no statistics for false alarms. Continuing research at the Acquabona site will include adding as many as 21 events to the analysis, producing more refined prediction and warning products from satellite imagery, and including thermodynamic analysis using atmospheric soundings from proximal stations.

Bibliography: Gregoretti, C. and Fontana, G.D., 2008. The triggering of debris flow due to channel-bed failure in some alpine headwater basins of the Dolomites: analysis of critical runoff, Hydrological Processes, 22, 2248-2263.

Holle, R.L. and Bennett, S.P., 1997. Lighting ground flashes associated with summer 1990 flash floods and streamflow in Tucson Arizona: An exploratory study, Monthly Weather Review, 125, 1526-1536.

Underwood, S.J. and Shultz, M.D., 2004. Patterns of cloud-to-ground lightning and convective rainfall associated with post-wildfire flash floods and debris flows in complex terrain of the western United States, Journal of Hydrometeorology, 5, 989-1003.