Is there a way to display the yaxis label in a Plotly/Dash heatmap row with NaNs only (empty heatmap trace)?

Question

I would like to show different heatmaps in different figure traces (Each heatmap with it's own colorscale, range etc.)

Therefore, I fill figure traces with individual heatmaps (go.Heatmap):

import numpy as np
import plotly.graph_objects as go
from plotly.subplots import make_subplots

fig = make_subplots(
        rows=3, cols=1,
        shared_xaxes=True,
    )

fig.add_trace(go.Heatmap(
                   z=[[15, 30, 40, -11, None],],
                   x=['Monday', 'Tuesday', 'Wednesday', 'Thursday', 'Friday'],
                   y=['Morning'],
                   coloraxis="coloraxis1"),
              row=1, col=1)

fig.add_trace(go.Heatmap(
                   z=[[30, 60, 1, -10, 20],],
                   x=['Monday', 'Tuesday', 'Wednesday', 'Thursday', 'Friday'],
                   y=[ 'Afternoon'],
                   name='Afternoon',
                   coloraxis="coloraxis2"),
              row=2, col=1)

fig.add_trace(go.Heatmap(
                   z=[[None, None, None, None, None],],
                   x=['Monday', 'Tuesday', 'Wednesday', 'Thursday', 'Friday'],
                   y=['Evening'],
                   name='Evening',
                   coloraxis="coloraxis3"),
              row=3, col=1)

fig.update_layout(
    
    height=400,
    coloraxis1={
        "colorbar_x": 1.0,
        "colorbar_y": 0.85,
        "colorbar_len": 0.4,
        "colorscale": 'plasma',
    },
    coloraxis2={
        "colorbar_x": 1.0,
        "colorbar_y": 0.52,
        "colorbar_len": 0.4,
        "colorscale": 'viridis',
    },
    coloraxis3={
        "colorbar_x": 1.0,
        "colorbar_y": 0.18,
        "colorbar_len": 0.4,
        "colorscale": 'delta',
    }
)

fig.show()

Problem

In case there is no data in one of the heatmaps (as shown the 'Evening' heatmap above) no yaxis label and no corresponding heatmaps are rendered.

I am looking for a way to show these labels also in case there are only NaNs

The only workaround I came up with was to fill dummy data into the heatmap. Is there another way to render the empty heatmap labels?