These are dark images (cap on). With severe stripes it can be very noticeable on uniform lighting and sharp contrast transitions as well but it is most prominent in low light scenarios.

The increase is likely caused by the buildup of heat in the sensor. The longer a sensor is energized, the warmer it will become and the noisier the output will be until the system eventually reaches equilibrium. Sometimes the sensor can be permanently damaged before equilibrium is reached. Many digital cameras used for photography as defined in the context of this community have safeguards in place that will shut the sensor off before it becomes hot enough to be damaged.

When there is no signal (light) striking the sensor, it's much easier to see noise, because that's all there is. The signal to noise ratio is basically 1:1, or 1.00. Even when there is some weak signal (low light scenario), noise is still much more noticeable than when there is strong light. As the light level increases the signal-to-noise ratio also increases.

Without knowing where exactly the camera's internal temperature sensor is located, it's hard to know if it is accurately reflecting the temperature of the sensor at the places where it matters: the photosites that capture photons and the circuitry that amplifies the signal from those photosites. This can be particularly a problem where a camera has multiple amplifiers that each process a portion of the sensor's output, usually in an interleaved pattern. Each amplifier circuit will not respond exactly the same to increased heat or increased light.