Ploting mean + SD's on a Temperature chart

That looks to me like a basic scatter chart with two data series. The "active warming" data series has positive error bars -- presumably taken from a column containing the standard deviation for each point. The "no active warming" data series has negative error bars -- again, I assume that the values for the error bars are taken from a column computing the standard deviation for each point.

Not knowing exactly what you have tried or what you are familiar and not familiar with, I would expect:

1) Compute average and standard deviation for each data series in the spreadsheet.
2) Select the time and average columns and insert a scatter chart.
2a) If Excel does not get the data series definitions correct, bring up the select data dialog and fix up the data series ranges.
3) Add error bars to both data series.
4a) Format the "active warming" series error bars to be only positive error bars and to use the values from the appropriate standard deviation column.
4b) Format the "no active warming" series error bars to be only negative error bars and to use the values from the appropriate standard deviation column.

What step do you have trouble with?

rows/columns should not matter -- as long as the chart series get defined correctly.

I cannot tell what you might be doing wrong from this side of the internet either. Can you upload a sample file (with anonymous data)?

For the average calculation, you are using SUM()/COUNT(), which is technically correct. Is there a reason you are avoiding the AVERAGE() function -- which should give the exact same result?

For the SD calculation, you are using the STDEV.P() function, which calculates the standard deviation for a population. The STDEV.S() calculates the standard deviation of a sample. I cannot say if you are understanding the subtle difference between sample standard deviation and population deviation and are using the correct one. Depending on what you are doing, it may be worth the effort to go back to your statistics text and review the difference and make sure you are using the correct one.

I tend to not use the ROUND() function for something like this, because I don't usually want to limit the precision of the final result. Only you can know if rounding to 2 decimal places is needed or wanted here.

The only mathematical question that I see is whether you should be computing population or sample standard deviation, and I cannot tell from here which is the case.