This work describes the creation of Progressive Glimmer, a progressive multidimensional scaling (MDS). It can quickly generate embeddings of data series with changing dimensions. For this purpose, the MDS algorithms Pivot MDS, CPU Glimmer, and the interpolation approach Majorizing Interpolation MDS are compared and evaluated. An MDS projects high-dimensional data onto data points with fewer dimensions, while keeping their pairwise distances equal. Progressive MDS are used to monitor and analyze high-dimensional data that is continuously updated, optimizing the embedding for the next update based on the result of the previous projection. While these often focus on adding new points to the data, Progressive Glimmer aims to handle changes in dimensions instead. Spatio-temporal data often doesn't change the amount of points, but encodes new arriving measurements as new dimensions in the data, or old dimensions are removed. When using MDS to analyze this type of data, Progressive Glimmer helps reusing the embedding for the next dataset and project it much quicker. Progressive Glimmer is a capable MDS, achieving the quality of CPU Glimmer with less variance in stress and up to seven times faster. The algorithm allows changes up to 20% of the original dimensions, although depending on the distribution of the data, the change can be much larger. The maximum amount of data that Progressive Glimmer could compute on our setup was 3 million points and 5800 dimensions.