Dual-polarization radar uses horizontal and vertical polarization to analyze the size, shape, number, and type of precipitation particles to accurately predict weather events. The output variables such as reflectivity, differential reflectivity, differential phase difference, non-differential phase difference, and cross-correlation coefficient are used to accurately observe various weather phenomena such as heavy rain, hail, and snow.
The importance of precipitation forecasting and weather observation
Weather phenomena such as heavy rain, hail, and heavy snowfall can lead to disasters, so it is important to predict precipitation and prepare for damage. In recent years, dual-polarization radar observations have enabled faster and more accurate weather observations, with precipitation information updated every 10 minutes. These advances in weather observation technology play an important role in reducing natural disasters. In particular, weather observation is also essential for agriculture, water management, urban planning, and other sectors. It provides important information for properly regulating the water supply needed for crop growth, preventing floods and droughts, and designing urban drainage systems.
How does dual-polarization radar work?
So how does dual-polarization radar observe weather events? Basically, weather radars send radio waves into the atmosphere and when they bounce off precipitation particles, they analyze the received waves and calculate several variables to analyze the precipitation particles. Dual-polarization radar also utilizes this principle, first determining the approximate size and number of precipitation particles through reflectivity, which is the comparison of the strength of the transmitted and received waves.
The radio waves transmitted and received by dual-polarization radar are composed of a horizontal wave, which oscillates in a direction perpendicular to the ground, and a vertical wave, which oscillates in a direction perpendicular to the ground. The reflectivity of each wave is called the horizontal reflectivity and the vertical reflectivity, and is measured in decibels per square meter (dBZ). The reflectivity used as an output variable of dual-polarization radar is the horizontal reflectivity, which is proportional to the size and number of precipitation particles present per unit volume of 1 m³. In general, drizzle with small and few precipitation particles has a value of 1 dBZ or less, while heavy rain with large and numerous precipitation particles has a value of 20 dBZ or more. However, in the case of hail, it is sometimes difficult to distinguish between the types of precipitation particles based on reflectance alone because the reflectance can appear similar to that of a downpour, even though the size and number of precipitation particles are different. Therefore, other output variables are needed to distinguish between them.
Utilizing different output variables
First, we can utilize differential reflectance to determine the size and shape of precipitation particles. Differential reflectivity is the horizontal reflectivity minus the vertical reflectivity, which is positive if the precipitation particles are longer horizontally and negative if they are longer vertically, and is measured in decibels (dB). For example, in a heavy downpour with large precipitation particles, the air resistance encountered by the raindrops as they fall causes them to spread out horizontally, resulting in a differential reflectance of more than 2 dB. On the other hand, if hail or snow has not melted and is composed of pure ice, even if the particles are large, they do not spread horizontally, and because they fall in a rotational motion, they are almost spherical and are often perceived by radar as having a differential reflectivity value of 0 dB. This allows you to distinguish between the weather phenomena of heavy rain and hail, even if they have similar reflectivity values. However, even drizzle with precipitation particles smaller than 0.3 mm encounters very little air resistance, so the differential reflectivity is often 0 dB because the precipitation particles remain spherical. Therefore, it is necessary to consider both reflectivity and differential reflectivity in combination to distinguish the types of precipitation particles.
On the other hand, knowing the type of precipitation particles such as rain or hail and the size of the precipitation particles is not enough to accurately estimate the number of precipitation particles per unit volume. Therefore, information about the number of precipitation particles is obtained through output variables called differential phase difference and non-differential phase difference. When radar waves hit a precipitation particle, the size and shape of the precipitation particle causes the horizontal and vertical polarization to progress at different rates. The phase of the two polarizations changes accordingly, and the cumulative value of the difference between these phases is the differential phase difference. The unit is degrees (°), and the phase difference is found by subtracting the phase of the vertical polarization from the phase of the horizontal polarization. The larger the cross-sectional diameter of the precipitation particle through which the radio waves pass, the larger the phase value, so just like differential reflectance, it has a positive value if the precipitation particle is longer horizontally and a negative value if it is longer vertically. Because the differential phase difference continues to accumulate along the direction of propagation, it is characterized by the fact that it can yield non-zero values even where no precipitation particles are present.
The rate of change of the differential phase difference at a given observation range is called the non-differential phase difference. If the differential phase difference is 0° at a point 5 kilometers away from the radar and 10° at a point 10 kilometers away, the differential phase difference between 5 and 10 kilometers is 1°/km, which is the change in differential phase difference of 10° divided by the round-trip distance of 10 kilometers. Unlike the differential phase difference, the non-differential phase difference only yields a non-zero value where precipitation particles are present, giving you a more accurate estimate of the number of precipitation particles in the particular section you want to observe.
The importance of cross-correlation coefficients
However, when two or more types of precipitation particles are mixed together, such as when snow melts and falls as both snow and rain, the output variable value may appear larger or smaller than the actual weather event, which can be confusing. The output variable to address this is the cross-correlation coefficient. The cross-correlation coefficient is a measure of the similarity of the horizontal and vertical polarization signals, with values closer to 1 the more similar the size and type of precipitation particles are. In general, the cross-correlation coefficient is high (0.97 or higher) when precipitation particles of the same type and similar size are distributed within the observation range during rain or snow. However, when there is a mixture of different types of precipitation particles, or when the precipitation particles are of the same type but vary in size, such as in a heavy downpour, the cross-correlation coefficient may be less than 0.97.
Thus, dual-polarization radar can utilize a variety of output variables to observe and analyze weather phenomena with greater precision. This allows us to prevent disasters and respond quickly to weather changes. Advances in weather observation technology will make our daily lives safer and more convenient.
