Algorithm for Improved QPE over Complex Terrain Using Cloud-to-Ground Lightning Occurrences

Carlos Minjarez-Sosa,Julio Waissman,David Adams,Christopher Castro

doi:10.3390/atmos10020085

Abstract

Lightning and deep convective precipitation have long been studied as closely linked variables, the former being viewed as a proxy, or estimator, of the latter. However, to date, no single methodology or algorithm exists for estimating lightning-derived precipitation in a gridded form. This paper, the third in a series, details the specific algorithm where convective rainfall was estimated with cloud-to-ground lightning occurrences from the U.S. National Lightning Detection Network (NLDN), for the North American Monsoon region. Specifically, the authors present the methodology employed in their previous studies to get this estimation, noise test, spatial and temporal neighbors and the algorithm of the Kalman filter for dynamically derived precipitation from lightning.

Highlights

Precipitation is one of the most important elements of the atmospheric hydrological cycle as well as the climate system [1,2]
The effect of using poor sensor coverage negatively affects Lightning-Precipitation Relationship (LPR), when non-covered regions were included, which is revealed in the decrease in explained variance R2 between CG lightning and precipitation from 0.39 to 0.23
By using 1-h accumulated lightning data, the study estimated convective precipitation, finding that about 67% of total precipitation for the 2005 season can be related with CG lightning

Summary

Introduction

Precipitation is one of the most important elements of the atmospheric hydrological cycle as well as the climate system [1,2]. Is necessary for accurate assessement of the atmospheric energy budget and hydrological cycle [7]. Under ideal conditions, are considered the most accurate point measurement of precipitation; for estimating QPE, geo-statistical methods are necessary and the accuracy of the estimation depends on gauge network density, which in most cases is not adequate [9,10,11,12]. On the other hand, are characterized as having high spatio-temporal resolution and coverage and have been considered the best operational option for estimation of rainfall [5,13,14]. Have limitations, such as poor coverage in mountainous regions, no unique Z-R relationship, and evaporation of rain from higher altitudes, among others [5,14,15,16]

Methods

Results

Discussion

Conclusion