
2
breakdown due to mass character. Also, the strength
of feedback would increase many times over [8], [6]. The
key parameter of an avalanche is its growth length, and
for real conditions in a cloud, it is quite large, so there
are few high energy particles in avalanche.
Naturally, a cell (a selected area of a thundercloud with
a directed electric field [29]) of a thundercloud cannot be
considered ideally homogeneous. The mass density of
hydrometeors in cumulonimbus clouds typically is up to
0.5 g/m3according to [30]. Approximations are usually
considered due to the small scale of the selected area.
Moreover, the presence of hydrometeors should be taken
into account.
Previously, hydrometeors were considered in streamer
physics. Studies have been carried out on the possible
role of corona discharges on ice hydrometeors [31]. A
class of hypotheses for the initiation of lightning and
sprites suggests that streamers are able to form around
the sharp tips of conducting objects (e.g., thundercloud
hydrometeors for lightning and ionospheric ionization
patches for sprites) placed in an electric field much
weaker than the value of the electric field at which the
electrons are accelerated rather than damped due to col-
lision [32]. That is, impurities in clouds were previously
considered as the cause of a change in the electric field,
and not as an effect due to a change in the mass fraction
of water in space, or, as the appearance of a dense inho-
mogeneity, a change in the direction of particle propaga-
tion and, as a consequence, a change in the total number
of formed particles.
In this work, we have shown that unexpectedly hy-
drometeors efficiently multiply runaway electrons. By
analogy with the interaction with air particles, runaway
electrons can interact with hydrometeors in thunder-
clouds. This results in a 20% reduction in avalanche
growth length for a realistic number of hydrometeors and
thus in the increase of the number of electrons.
In this article, the influence of the presence of hydrom-
eteors in thunderclouds is studied. Also, the influence of
the atomic composition of the substance is considered
specifically, without taking into account the change in
the electric field by the particles of hydrometeors. This
article presents the analysis of RREAs development in
the cloud, based on numerical modeling. Describes two
different modeling options taking into account hydrome-
teors. The second section ”Modeling” describes the mod-
eling of energetic particles behaviour in the media of the
cloud modeled in two different ways: model with volu-
metric hydrometeors and model with modified material.
Section 2.2 presents the simulation results with volumet-
ric hydrometeors. A comparison is made of the number of
produced electrons with an increase in the mass fraction
of hydrometeors and one radius, as well as one mass frac-
tion and different radius.A comparison of the spectrum
at the exit of a cell with and without a small number
of volumetric hydrometeors, and a simulation with one
hydrometeor is also presented. Section 2.3 presents the
simulation results with modified material. The results in
the section prove that the effect obtained is not a con-
sequence of a change in the nuclear composition of the
substance (adding an ice component).
III. MODELING
A. General properties
Numerical modeling is one of main instruments in at-
mospheric physics, crucial in conditions of limited mea-
surement data, and useful for analyzing complex mecha-
nisms verifying the results of other studies. In present
study we carried out Monte Carlo simulation using
GEANT4 version 4.10.06.p01. GEANT4 is widely known
and used for the problems of simulating the passage of
particles through matter. Its fields of application include
high energy physics, nuclear and accelerator physics, as
well as medical and space research [33].
We used the Physics list
G4EmStandardPhysics option4. It should give ac-
curate results in electromagnetic physics simulations.
We developed two approaches to modeling the cloud
media with hydrometeors: model with volumetric hy-
drometeors and model with modified material. Within
both approaches, the modeling volume was a cube with
a side of 200 m. The cube initially consists of air with
a density of 0.414 mg/cm3, which corresponds to the
density of air at a height of 10 km. Next, depending
on the chosen approach, we add hydrometeors. In the
cube the uniform electric field with strength 200 kV/m
is applied. The direction is chosen so that the electrons
are accelerated in the direction of launch. The minimum
step for simulation was chosen as fMinStep = 0.01 mm.
An electron with an energy of 5 MeV is launched. We
fix the born particles in the entire volume using the
SensitiveDetector (the sensitive detector class in Geant4
has the task of creating hits (deposits of energy) each
time a track traverses a sensitive volume and loses some
energy).
B. Model with volumetric hydrometeors
In this simulation, hydrometeors with a radius of 1
cm were used, which is bigger than a real hydrometeor.
However, modeling smaller balls leads to an increase in
their number at the same mass fraction, which compli-
cates the calculations. In order to justify the use of parti-
cles of this size, the following additional simulations were
carried out.
A cube with a side of 50 meters was taken, since in
this case one can limit oneself to measuring the number
of electrons, and not the growth length of the avalanche.
For exact values of the growth length, one should take
the length of several the growth length of an avalanche
so that it has time to stabilize. For us, in order to
model many small-radius hydrometeoroids for a given