Atmospheric Radar Research Center

When establishing the operating parameters for a weather radar, one would ideally like to maximize the detectability of the received signal while minimizing the sampling volume. Conventionally, these two objectives are at odds with one another. For a given available peak transmit power, signal detectability is maximized by increasing the length of the transmit pulse, while improving the range resolution requires that the transmit pulse be decreased.
This assumes, however, that the product of transmit pulse length and the amount of bandwidth is unity. In reality the bandwidth can exceed this criterion by coding the transmit pulse in either frequency or phase. The excess bandwidth is used to improve the range resolution. The use of pulse coding is a long-standing and established practice for military and aircraft tracking radars, which focus on point scattering targets.

Later, pulse coding for observations of distributed atmospheric targets was introduced for both weather [Fetter, 1970; Austin, 1974] and profiling radars [Farley, 1969; Schmidt et al., 1979].
Until recently, however, pulse compression technology has only been routinely used within the radar profiling community. The reasons for this are outlined in Keeler and Passarelli [1990]:

1. adequate resolution with conventional weather radars to achieve scientific goals
2. limited dwell times due to mechanical scanning rates
3. the effects of range time sidelobes

However, recent developments have led to a renewed interest in this topic for meteorological radars [Mudukutore et al., 1998; Ohora and Bech, 2005].

The current interest in pulse compression is particularly well articulated by the design considerations for MPAR, which include the integration of low-power solid-state transmitters, digital beamforming functionality, and advanced signal processing capabilities.
A simple metric to quantify the effectiveness of a particular pulse coding scheme is the so-called time-bandwidth product, BT. Here B and T are the bandwidth and duration of transmitted pulse, respectively. For example, the range resolution of a conventional uncoded (BT = 1) 20-μs pulse has a range resolution of 3 km; however, using a pulse compression scheme with a BT of 20 the range resolution is reduced to 150 m. Ohora and Bech [2005] claim that BT values of 100 to 200 can be achieved using meteorological radars. They present data from two weather radars, which were operated with 8 kW of peak transmit power. It is shown that through frequency modulated pulse compression techniques these radars were able to achieve range resolutions equal to or better than a more traditional weather radar having a peak transmit power of 250 kW. This corresponds to a BT of about 31.

Here it is proposed to investigate the impacts of pulse compression on MPAR using both the phase and frequency modulation. Whereas Mudukutore et al. [1998] discuss the benefits of pulse compression though binary phase coding, Ohora and Bech [2005] maintain that frequency modulation offers several advantages over phase modulation such as improved integrated sidelobe level (ISL) suppression. Sidelobe suppression is particularly relevant when making meteorological observations due to the nature of the scattering targets. Therefore, it is important to examine various pulse compression methods in a controlled manner. This will be primarily achieved through the use of numerical simulations with the proposed radar simulator. Various virtual radar experiments will be designed and tested within realistic weather scenarios for a variety of waveform designs. In this way it will be able to ascertain which method is optimally suited under various meteorological conditions.

In addition to running simulations we plan to work together with the researchers at NSSL to explore the possibilities of implementing pulse compression on PAR. If deemed feasible, then a series of pulse compression experiments will be conducted using PAR and the performance enhancements will be compared against theoretical predictions.