VNP46A3 off-nadir viewing angle

[nikos_geography]

Hi, I'm using the VNP46A3 data. My task is to downscale the near-nadir (NN) and off-nadir (ON) products and measure the impact of the viewing angle (VA) on the blooming effect (or PSF) from a geometric perspective. Let me explain.

If I project a viewing cone from nadir the footprint (support) is circular and has minimum area (given fixed viewing height).

If I project the same viewing cone at an angle, off-nadir, then the support is an ellipse and it has a larger area. The greater the angle off-nadir, the more pronounced the ellipse and the larger the area.

The areal increase can be calculated from geometry (my problem).

I am interested to test if the predicted increase in the PSF is equal to this known geometric effect.

What I have done so far is the following:

Assuming that the PSF width (measured in pixels) of the NN product is 1 pixel and the PSF of the ON product is 1.3 pixels. These two values are what I predicted during the downscaling.

To test the predicted value of the ON product, I did:

1/cos(40) ~1.34 which is close to what I predicted. The number 40 is the viewing angle of the sensor (in degrees) and 1 is the PSF width of the NN product.

Here are my questions:

1) is this geometric solution valid? Does it make sense?
2) where can I find the actual degrees of the viewing angle in each NTL image?

Many thanks.

[LAADS-EDL_UserServices_M]

The monthly composite product (VNP46A3) is created using observations from varying view angles collected throughout the month. As a result, a single view angle cannot be isolated for the off-nadir composite. Pixel-level view angle information is only available in the at-sensor top-of-atmosphere product (VNP46A1) under the layer “Sensor_Zenith.” For more details, please refer to the user guide.
https://viirsland.gsfc.nasa.gov/PDF/BlackMarbleUserGuide_v1.2_20220916.pdf

[nikos_geography]

Thank you for your response. Let's say I use one daily image and not the monthly composite, is the equation I posted correct?

d/cos(θ)

where cos(θ) is the sensor's VA and d is the PSF at the (near) nadir.

[LAADS-EDL_UserServices_M]

The spatial resolution of DNB remains stable across varying view angle since each pixel is aggregated from several detectors. The following document might be helpful:

(https://ladsweb.modaps.eosdis.nasa.gov/api/v2/content/archives/Document%20Archive/Science%20Data%20Product%20Documentation/Product%20Generation%20Algorithms/NASARevisedVIIRSGeolocationATBD2014.pdf).