Context. Autoscaling is a technique for dynamically adjusting the number of pods in a Kubernetes cluster based on resource utilization or custom metrics. Problem. However, autoscaling decisions are often opaque and hard to understand for developers and operators of self-adaptive microservice applications, thus verifying and debugging their behavior is difficult and time-consuming. Objective. This thesis aims to improve the insight into horizontal pod autoscaling (HPA) decisions by developing a concept that presents data of scaling decisions made by HPA in a visual and interactive way. Method. For this purpose, previous work on this topic was analysed. Questions and requirements for a system to explain this autoscaling behaviour were elicited. Result. The concept was implemented as a web-based dashboard that integrates with Kubernetes clusters to monitor the scaling behaviour of a self-adaptive application and displays various metrics and components related to the scaling behaviour of the monitored system. Conclusion. The dashboard was evaluated with an expert survey. The results show that the dashboard can help users gain more awareness and understanding of autoscaling behaviour. In particular, fewer different tools are needed to gain insight into scaling behaviour.