In remote sensing, understanding the application and necessity of ground truth data is crucial. Let's explore the role of ground truth data in the context of the provided options:

1. Calibration of sensor: Calibration is the process of configuring a sensor to ensure accurate results under specific conditions. Although ground truth data can be used in the context of checking calibration efficiency, it's primarily more impactful in the analysis stages than in calibration itself.
2. Analysis of remotely sensed data: Ground truth data plays a critical role here. It acts as a benchmark to validate and interpret the data acquired from remote sensing. Hence, ground truth is paramount in ensuring that the analysis reflects real-world conditions accurately.
3. Accuracy assessment: Ground truth data is essential for assessing the accuracy of remotely sensed data. It helps in verifying whether the classified images or data collected represent what is on the ground, ensuring the reliability of remote sensing outputs.

Based on these explanations, we can conclude:

• a only: Ground truth data is not primarily used for only calibration, and this statement alone does not justify the extensive role of ground truth.
• b and c only: This option is correct as it encompasses the major roles where ground truth data is critically involved, i.e., in analysis and accuracy assessment of remotely sensed data.
• a, b and c: This would imply an equally prominent role of ground truth in calibration as well, which is misleading.
• a and b only: This neglects the crucial aspect of accuracy assessment, which ground truth is vital for.

Therefore, the correct answer is b and c only.