weiji14
commented
1 year ago Hi @Bbbigcountry, thanks for your question!
From my understanding, there should only be one data point per cycle for this anomaly. However, I have noticed that the crossover points data appears to be too time-intensive.
Each blue line on the plot represents a crossover point within the subglacial lake boundary. The top left corner of the plot shows a map of the subglacial lake, with the light blue dash-dot line being the lake boundary, and the diamond points inside the polygon being the crossover points. Those diamond elevation points are coloured so that they are darker blue if the elevation anomaly is greater. You will notice clusters made up of 9 points, which is when 3 ICESat-2 ATL11 pair tracks crossover with 3 other ICESat-2 ATL11 pair tracks (3x3=9).
So, there would not be just one data point for each crossover point in time, but several. In that regard, it would appear to be 'time-intensive', because you could have about 9 points at each crossover cluster, potentially 27 or 36 if there are 3 or 4 such crossover clusters in one crossing. This will be a lot more data than single elevation anomalies others usually put out for a single subglacial lake. If you're interested in the 91 day rolling mean of elevation anomalies, it is the dashed black line.
You can find more details in Chapter 4 of my thesis at page 86 https://doi.org/10.26686/wgtn.14956062.v1 (LaTeX is at https://github.com/weiji14/phdthesis/blob/0864ba94c9a6abbd1548ae7e2ce10c7089f0b4d0/chapters/04_chapter.tex#L587-L628) if you're interested. The code to generate the plot is at https://github.com/weiji14/deepicedrain/blob/v0.4.2/atlxi_xover.py#L284-L320 and https://github.com/weiji14/deepicedrain/blob/v0.4.2/deepicedrain/vizplots.py#L302-L467, and the data should be inside https://github.com/weiji14/deepicedrain/releases/download/v0.4.2/df_dhdt_whillans_upstream.parquet. Let me know if you need extra help figuring out.
Bbbigcountry
Hello, I have a question regarding the line chart in readme. Specifically, I would like to inquire about the elevation anomaly at crossover points. From my understanding, there should only be one data point per cycle for this anomaly. However, I have noticed that the crossover points data appears to be too time-intensive.