Polarimetric Data Analysis with IDL

This blog, written by Ranier M.E. Illing is the third blog of a series writing by our IDL Fellows. The IDL Fellows program is our way of supporting passionate retired IDL users who may need support to continue their work with IDL. There is continual innovation behind IDL, and this program is one of the many ways we hear about the new innovative ways people are using it on a regular basis. If you are retired and interested in becoming an IDL Fellow and sharing your research through a blog post, feel free to reach out to me to see if you qualify for the program.

When we look at the world around us, most of the information our eyes capture comes from the intensity of light. We perceive different colors based on the varying intensities in spectral bands, helping us understand the characteristics of objects and their surroundings. However, our eyes are not particularly adept at detecting another critical quality of light – polarization.

Light is more than just color and intensity. It is a wave that oscillates perpendicular to its direction. Polarization measures the coherence of this wave in time – essentially whether the wave oscillates in a fixed plane, spirals around its path, or something in between. To fully describe the properties of a light wave, scientist use four key parameters.

S0 – the total intensity of the wave
S1 – the amount of intensity oscillation in a given transverse direction
S2 – the amount of intensity oscillation at 45° to that direction
S3 – the amount of intensity oscillation around the line of sight

Linear polarization, where the wave oscillates in a fixed transverse plane, is the most common form. For instance, polarized sunglasses filter out all light that doesn’t oscillate in a horizontal plane, reducing glare and improving visual clarity.

A Closer Look at Polarization with IDL®

For our exploration of polarized data, we acquired field data images using the Ball TTP dual liquid crystal polarimeter. The raw images were processed and corrected using the PolarQL analysis system written in IDL. IDL enables us to analyze the processed images and generate a variety of polarimetric products, offering deeper insights into the data.

One particular dataset stood out. It focuses on the elimination of haze using only polarized light measurements. The study took place on the grassy knoll in front of the Ball BCH, with the TTP set up to capture imagery of Eldorado Canyon.

Discovering Hidden Layers with Linear Polarization

Figure 1 shows a sample of the linear polarized intensity gathered (calculated as S_linear=√(S_1^2+S_2^2 ) ). The vibrant color display reveals the different layers of mountain ranges, each differentiated by their linearly polarized scattered path radiance. This unique visualization allows us to see through the haze, distinguishing the various layers of landscape that might otherwise blend together.

To further explore this data, we used the interactive segmentation tool available in PolarQL.pro version 2.1. This tool allows any pair of data images to be plotted against each other as a 2D histogram. The data available for analysis include the four raw images (S0, S1, S2, S3), linearly polarized intensity, fractional polarizations, ellipticity, and more. Users can interactively select a region of interest using a box cursor, which highlights the corresponding pixels in a “show” image rendered in a red color table. The image show can be any of the available raw or processed images.

Figure 1. The first image of the two below shows the linear polarized intensity, with the vibrant color showing the layers of mountain ranges, each differentiated by their linearly polarized scattered path radiance.

Figure 2. Eldorado Canyon data analyzed using the segmentation tool in IDL. Features in the S1-S2 histogram plot pointer box are highlighted in the S0 image as green pixels.

Figure 2 shows the clear segmentation of the image by its linear polarized flux. The green section corresponds to the pixels in the cursor box connected by the arrow, clearly highlighting the segmentation of the image based on its polarization characteristics. Notice the shadowing effect – the farther mountain range’s polarization is overshadowed by the nearer mountain, indicating a significantly different polarization between the two.

Beyond Fractional Polarization

While fractional polarization is important, it can sometimes obscure clear segmentation seen in linearly polarized flux. This is because fractional polarization involves division by S0, introducing intensity features with polarization information. For example, a low polarized intensity feature that can be clearly distinguished from a high polarized intensity feature, such as the sky and the front mountains in Figure 2. However, after dividing the total intensity, the fractional polarization might become similar for both features, making it harder to distinguish.

Instead of relying solely on fraction polarization, the “most polarimetric” information can be found in the polarization angle (or more directly, S2/S1). This ratio focuses solely on the polarization of light, eliminating the need for absolute intensity calibration and providing a more accurate representation of the underlying polarimetric properties.

Polarimetric data analysis offers a powerful lens through which we can uncover hidden details in our environment, revealing insights that go beyond what is visible to the naked eye. By leveraging tools like IDL and the PolarQL analysis system, we can explore the nuances of light's polarization, enabling clearer interpretations of complex datasets.