what do you mean with strong accuracy?

you should also mention in the title at which scenario you are looking here

Nice start, unfortunately the climate action of today only pays off when the Trump is out of office again, so he probably will not bother. But this is how the system works. Still a good question you raise.

dramatic sentence

Reference? Where did to took this information from?

what do you mean with significant? (typical ChatGPT word)

when you introduce a new indicator, you should shortly mention what it is showing

Reference for this?

this is not true from the plot, in 2100 Mexico city has an equal number than phoenix, but less than new york

should be bold ;)

should be in a new line and bold ;)

For a real comparison you should only look at the temperature increase at glacier locations and not the entire country. But this is a bit advanced with the provided dashboard data.

This is the only important sentence for your analysis. You should not only summarize what you found in your reference, you should also set it in contrast to your findings (e.g. move this last sentence to your analysis above)

Percentage of what

I do not understand what you want to say with this sentence

Do you have an idea why?

Great spotting this. I will contact the dashboard developers and let them know.

have you looked for an explanation for this?

this images are also not workin

this images are not working

this is not a good reference, you should link here to an actual paper, instead of the entire website

better to including the references in the text when they fit (e.g. comparing your results with their results). But an additional list of all used references at the end is great!

You have many spelling mistakes in this paragraph. Maybe check your writing with a program like grammerly

mention in the title which scenario you are showing here

add here also the name of the variable shown

Why is this a failed try, the plot below does look ok

If you encounter problems like this, you should reach out to us, so we can help you!

you probably could also find another reference, which is accessible to everyone

this is a repetition of the sentence before

their is something missing

two

correlation

unit is missing

again a good interpretation of the graph

Good interpretation what the graph is showing

good usage of references. However, you should provide a list at the end, where all are listed.

You have not answered this questions

You selected only one scenario!

you have not included any custom plots

Providing a reference without saying what is in their is not enough.

Again Reference

Why you are not linking to the graphs above?

Reference for this?

Reference for this?

you answered those question already during your analysis. You do not need to repeat them here again

you should have provided the references within your analysis, whenever it fits (e.g. comparing your findings to a reference, ...)

What do you mean with significantly? (typical ChatGPT word)

'changes in fire season length' are the same unitl 2040, 'heatwave days' are the same unitl 2050

where did you download this 'absolute'-data. I only could find 'present-day'

when i check the graphic on the dashboard I see only 33 days in 2100... Where did you took the data from?

in which year? (in 2100?)

I assume you do this analysis for the year 2100. You should mention this here

plots in this section are not working. I also checked the notebook and they are also not working their. You always should check this before submission

same as for indicators

it would be enough to shortly summerize the indicators in your own words, no need to copy the text from the dashboard

well structured analysis, but your whole question 5 could be shortened, as you repeat yourself.

why 'once again'

you should provide this references inside your analysis, where it fits (e.g. when you compare your findings to the references or if you cite something from the studies)

what do you mean with significant? (typical ChatGPT word)

It should be explained somewhere before, what this two parameter are, and how you came up with this formulation

did you came up with this risk score, or is their a reference you took it from? If you came up with it, you should explain it in the text, not only in the code. Or you should give the reference for further details

Probably you could think of a nicer graphic illustrating this (e.g. two plots below each other, upper plot showing x-axis time and y-axis temperature, lower plot showing x-axis time and y-axis change of days with heat stress.

for strengthen this argument you could calculate the increase in days per 10 years (e.g. from 2080 to 2100 we see an increase of around 3 days per 10 years, compared to earlier periods with only 2 days increase in 10 years)

add something like: ", shown in red",

from the graph it is closer to 10, maybe 11

You could use a discrete colorbar here, instead of a continues, as you only want to colercode specific years.

What do you want to say with this? You just give again the definition of heat stress.

reference?

good conclusion of the shown graph

how do you define a real threat?

where is this number coming from? reference?

well structured explanation of the selected scenarios

typo

well formulated research question

you should use delayed_data here

Why do we see the pattern for stabilization and delayed action, with the strong drop of the curve before 0.5?

typo

thinning rate of 0 could mean glaciers are in a new equilibrium -> do not melt further, or they are completely gone

Good conclusions

the reference should be moved to the next sentence

This is only true for SSP1-1.9, not for SSP5-3.4-OS where temperatures are declining after 2060. You should link here to the plot where it is shown.

You have not included what scenario is shown in the plot. You have not explained where the minimum and maximum volume are coming from.

you have not answered any guiding questions

you could put numbers here from the dashboard data

How does this link to the dashboard data?

But it saves at least 18% of todays volume, compared to 0%

From where do you draw this conclusion. Reference?

This whole paragraph reads like a quick chat-gpt output, without any prove reading or adapting

you should link to the plots. For SSP1-1.9 at their will be 18% of the volume left...

what does this mean? reference?

You can not use the data of the dashboard to answer this. For this you would need data for specific glaciers of individual ski resorts and how they change with different scenarios.

No caption for temperature and sst anomalies, no references for comparing your findings with other sources

Which plots

About what anomaly are you talking about here (temperature or sst) ?

La

Here also, would be better to use a divergent colormap, for easier distinguishing between positive and negative differences https://matplotlib.org/stable/users/explain/colors/colormaps.html#diverging

Would be better to use the same colormap as for the temperature, to make it easier to distinguish between positive and negative values

It is better practice to select the year explicitly here, not just using an index, because with an index it is easier to make typos

Link?

Missing in the answer

would be nice to mention the process at play (photosynthesis)

You should have used the weighting here!

Great that you found that the given number of 55 was wrong. They recently changed the file format, sorry for that.

And what are the patterns you see?

Yes, but why have you filtered those out?

Why do you filter out negative precipitation differences (where we see more precipitation in la nina case compared to el nino?

Not a good colorscale for looking at differences. Better would be a divergent colormap, where you can easily distinguish between positive and negative values https://matplotlib.org/stable/users/explain/colors/colormaps.html#diverging

Where do you see this?

This is not a anomaly this is a difference. A anomaly is calculated against average conditions (e.g. the mean over the whole period). See third session of practicals

Similar as above, why have you selected two years here?

Why have you picked two years here?

You should have weighted the temperature according to latitude (see first practical session)

Great that you find out, it was a mistake from my side as they changed the file-format of the data recently.

Description is completely missing

What is this?

figures too small, I can not read anything without zooming. Better to create more figures, e.g. one per variable.

Which periods have you picked?

Why are you skip lines at the end of the file? By doing so you miss most recent datapoints

You correctly identified our mistake of using 38 instead of 55. Great

From where did you take this value (reference)? Ok, just saw in before the plot you used 1980-1985, but the pre-industrial era is way before 1980 (around 1800, see this link https://www.climate.gov/news-features/understanding-climate/climate-change-atmospheric-carbon-dioxide)

You are missing the latitudinal-weighting of temperature (see first session of practicals)

wrong years, we asked for 2016 - 2021

1980 to 1985 is not the per-industrial era (see comment above), but value is correctly calculated, with an interesting slope approach :)

only partly correct, as the driving process for the seasonality is photosynthesis of the plants growing on this landmass

You should have used your plots for discussing the differences. You just summarized information from an unknown source (at least you do not mention your source). Same for precipitation and air temperature.

Why are you skip lines at the end of the file? By doing so you miss the most recent data.

Link is nice, but you should also summed up the content in one to two sentences.

link alone is not enough, you should also summaries in one to two sentences

do not need the +1 here (its 6 years but we calculate the mean here for 5 differences)

The temperature should have been weighted according to the latitude (see first sessions of practicals).

Great, you correctly identified our mistake of using 55 here. The problem is that they changed the file-format.

Where do you see this? Globally? Are their regional differences? No comparison to other sources using references in the whole exercise!

Again, these maps are not well suited to make such an analysis. Better use anomaly maps! What do you see for La Nina?

Where on the globe do you see this? Is it everywhere the same?

It is hard to make such an analysis by looking at two plots with total values. Better use anomaly maps below for it.

And what do you see during La Nina?

This color levels are not well selected and most of the globe has the same color and you can not see any differences, it is only good for looking at the pacific...

reference?

no unit in colorbar-label

no title, what is shown on this plot? Same for the other plots.

Which sources?

From where did you get this years?

temperature should have been weighted by latitude (see first practical session)

I do not understand what you tried to say here. References?

wrong calculation, you should have taken the difference from the annual average of 1985 and 1980 and divide by 5, e.g.

dfa = df['average'].groupby(df.index.year).mean()

conc1 = (dfa.loc[1985] - dfa.loc[1980]) / 5

reference?

This was a mistake from our side and should be 38, but good workaround with set_axis.

Great interpretation! Could have compared your findings to others by adding references.

influencing

could have used nicer labeling for the colorbars, including units

This is not a zonal anomaly, just a anomaly. For a zonal mean the reference (or baseline) is the 'longitudinal-mean', see definition of the third practical session. The same for all following plots.

For the temperature you should have used weight according to latitude (see first session of practicals)

Important to mention her is that northern hemisphere has more land mass and therefore we see in the northern spring and summer months the higher uptake. If south and north hemisphere would be equal in land distribution we would not see this cycle in the way we do.

annual average CO2 Concentration

Mistake from our side, this should be changed to 38, with this you would not need to define the columns. But great solution! This will not effect the grading.

reference? just from the map south US looks not much different than north US

Not only copying a slide from the lecture, but discussing your findings with more references would be nice

same as above

same as above

difference

same as above

You could have been more precise here, e.g. where exactly do you see the warmest/coldes temperatures? Directly at the cost or more in the middle of the ocean?

I do not understand where to find the this information

You are not showing a zonal anomaly here (see third practicals lesson for definition). You actually showing the difference between an El Nino year and a La Nina year. And for a zonal anomaly you first need to calculate a mean value along the longitude and afterwards subtract it from the total signal.

You could also have chosen all years of a 30 year time period, but also fine to pick one year. But when using one year it is not an anomaly it is called difference. For anomaly see for example question from practicals: compute the temperature anomaly for the year 1997 with respect to the 1979-2018 average.

difference

No zonal anomaly, see comment above

No zonal anomaly, see comment above

This is the value from 2014 not 2016, see comment above

You used the wrong value for twenties_increase (instead of 2016 you used the value from 2014). To avoid such things should not do copy and past values, instead us selection of pandas: e.g.

twenties_increase = (co2_year_av.average.loc['2021-12-31'] - co2_year_av.average.loc['2016-12-31']) / (2021 - 2016)

minor comment: normally the variable names indicate which data they represent, ds stands for xarray.DataSet. However, in this case we using pandas and what we get is a pandas.DataFrame, so normally you should use df for this (e.g. df4 in our case). This has no influence on the grading, just to let you know

My mistake this should be 38, as they changed the file format from last your to this year of co2_mm_gl.csv. But great solution with setting the axis by yourself!

Error should not be happen in the final document.

You plotted the difference of the two events, but you should have plotted anomaly maps for the two events, which are always against a reference (e.g. 30 year period or year with neutral phase). See third lesson of practicals.

No discussion of sea surface temperature

Hard to easily see this because of the used colormap. A divergent colormap with a clear distinction between positive and negative values would be better (see comments below).

same comments as for precipitation plot above

When yo showing differences you should have used a divergent colormap https://matplotlib.org/stable/users/explain/colors/colormaps.html#diverging, to make it easier to identify regions with precipitation increase and decrease (e.g. the same colormap as for temperature). Why have you used unevenly spaced levels (negative values have a difference of 5, where positive values have a difference of 2.5)? Their are reasons to do something like this, but it make it often harder to interpret the results intuitively.

this is not true, you only should have stated here east Russia and Alaska, as their are also regions of the higher latitudes with colder temperatures

Also mention if it is warmer or colder

Also mention if it is more or less percipitation

same as above

The temperature increase is not evenly over the complete regions you mention, their are also parts with colder temperatures. When presenting results you must be more precise (like in the discussion below of Australia alone)!

This caption is not at the right place

Try to avoid naming the total dataset the same as a variable inside, this can get quite confusing if someone else is reading you code. For a dataset it is common to indicate the datatype in the variable name (e.g. ds_sst stands for dataset_sst). Only a recommondation

Be careful with your wording. In science, when you use the word significant it means you conducted some sort of statistical test to show that the results you are presenting are 'statistically significant'.

same as above

never use python functions like sum as variable, because you override their functionality (e.g. before you defined sum you could have used sum([1,2])). Normally those functions are highlighted green in jupyter.

delay

Calculations are correct! But would have been more effective if you just calculate the difference of the first and the last year. Here as an example for 1980-1985:

dfa = df.resample('AS').mean()

conc_1980_1985 = (dfa.average.loc['1985-01-01'] - dfa.average.loc['1980-01-01']) / 5

vs normally means x-axis vs y-axis, so by just reading the heading I would assume CO2 on the x-axis and temperature on the y-axis. Maybe better would be 'CO2 concentration and temperature'

You could also just use weighted_mean.resample(time='AS').mean().plot(), without the need of defining 'jahre'

you plotted the average monthly concentration, but the 'raw' timeseries would have been enough (e.g. df.average.plot())

Sorry, they changed the file format, but you correctly replaced it with 38. But you could have left 'parse_dates' and 'index_col' as this combines years and months into a combined timestamp, which is easier to work with.