Model agreement and trend analysis data associated to the publication: "Impact of climate change on site characteristics of eight major astronomical observatories using high-resolution global climate projections until 2050"

This dataset is associated with the following publication:

Haslebacher, C., Demory, M.-E., Demory, B.-O., Sarazin, M., and Vidale, P. L., “Impact of climate change on site characteristics of eight major astronomical observatories using high-resolution global climate projections until 2050. Projected increase in temperature and humidity leads to poorer astronomical observing conditions”, Astronomy and Astrophysics, vol. 665, 2022. doi:10.1051/0004-6361/202142493.

In the folder 'model_agreement', there are pickle files from which a python dictionary can be extracted with:

with open('mypklfile.pkl', 'rb') as myfile:
dload = pickle.load(myfile)

Pickle files ending with '_d_obs_ERA5.pkl' contain in situ data and ERA5 data. Pickle files ending with 'd_model.pkl' contain PRIMAVERA model data. A few explanations:
- 'ds_sel': contains monthly timeseries of selected intersecting data
- 'ds_taylor': contains data used for the Taylor diagram (Figs. 4-10)
- 'ds_mean_month': contains seasonal cycle for plotting (Figs. 4-10)
- 'ds_mean_year': contains yearly timeseries for plotting (Figs. 4-10) 

The subfolder 'median_nc_u_v_t' contains NETCDF files with the median and interquartile range of the wind speed in u and v direction, the temperature and geopotential height. This was used for Figs. G1-G8 and to calculate the refractive index structure constant Cn2.

The subfolder 'skill_score_classification' contains csv files with the sorted skill score classifications. The column headers are: model_name, skill score, correlation coefficient, standard deviation, centred root mean square error.

The folder 'trend_analysis' contains for each variable csv files of ERA5 and PRIMAVERA monthly time series used for trend analysis, pdf files of analysis summaries, csv files of Bayesian analysis results and png files of longitude-latitude maps of trends (analysed with linear regression). Additionally, there is a csv file of averaged in situ pressures.

Code that generated and used this data is available on github: https://github.com/CarolineHaslebacher/Astroclimate-future-project