I am close to solve this , can guide me so I can contribute..

I am close to solve this , can guide me so I can contribute..

Well, you need to understand the code in python/grass/jupyter/baseseriesmap.py and timeseriesmap.py to be able to solve this. I would start with trying to reproduce the problem. But if you are close to solving this, I am not sure what guidance specifically do you need?

@petrasovaa I only have two questions Should layer removal (to prevent stale layers) happen in TimeSeriesMap._render() or baseseriesmap.py’s _update_figure()? Are there existing functions to track/remove layers by timestep, or do I need to implement new logic.
Should TimeSeriesMap clear all layers between timesteps, or only those explicitly added via add_raster_series/add_vector_series? Is the root cause incomplete cleanup or incorrect rendering order.

@petrasovaa I only have two questions Should layer removal (to prevent stale layers) happen in TimeSeriesMap._render() or baseseriesmap.py’s _update_figure()? Are there existing functions to track/remove layers by timestep, or do I need to implement new logic. Should TimeSeriesMap clear all layers between timesteps, or only those explicitly added via add_raster_series/add_vector_series? Is the root cause incomplete cleanup or incorrect rendering order.

I don't know what is the root cause, I didn't investigate. What do you mean with stale layers? I don't see why you need to remove any layers.

@petrasovaa I think layer cleanup should happen in TimeSeriesMap._render() so we don’t accidentally break other parts of baseseriesmap.py. The goal is to only remove layers added by add_raster_series or add_vector_series (like temporary rainfall maps) while leaving static layers (e.g., a base map) untouched. The bug seems to happen because those temporary layers aren’t fully cleaned up between timesteps, not because of rendering order.

Here’s what I’ve tried: Reproduced the issue- When rendering timesteps, old layers stick around and overlap new ones.
Added a fix- I modified TimeSeriesMap to keep track of temporary layers and remove them before drawing new ones,
about to test it locally With the NC dataset, layers now refresh correctly without leftovers.
Does this approach make sense? or is there something I’ve missed?

Why don't you make a draft PR and I will look at it, without seeing the code I can't tell.

@petrasovaa Yes, I'll do it shortly.