Problems

You might also like

Download as docx, pdf, or txt
Download as docx, pdf, or txt
You are on page 1of 3

Quiz 2

Read the document carefully

Problems:

Group 1: Download the NCEP-NCAR data for Rainfall. Extract data for the months June,
July, August and September for all the years given in the file only for India. (Plot wherever
necessary)

Take the spatial average over the country. Fit all the distribution over the data and find which
distribution fits the best. Show all the plots for all the distributions.

Coordinates for India- 66o E – 100o E and 6o N – 38o N

Variables to be downloaded: Precipitation rate - Monthly mean dataset from link


https://www.esrl.noaa.gov/psd/data/gridded/data.ncep.reanalysis.derived.surfaceflux.html

Group 2: Download the NCEP-NCAR data for variables Evaporation, Total Precipitation
and Runoff. Find relationship between Evaporation and Total Precipitation with Runoff for
the month of August for India. Take the spatial average over India. What do you infer from
your analysis? (Plot wherever necessary)

Coordinates for India- 66o E – 100o E and 6o N – 38o N

Variables to be downloaded: Precipitation rate (Rainfall), Latent heat net flux (Evaporation)
and Water Runoff - Monthly mean dataset from link
https://www.esrl.noaa.gov/psd/data/gridded/data.ncep.reanalysis.derived.surfaceflux.html

Group 3: It is found that there is some relationship between temperature over a sea surface
region and All India Monsoon Rainfall (i.e. The sum of rainfall over all the grids over Indian
region). But similar is not the case with the relationship between this sea surface region and
each grid of India.

Find the correlation between the temperature over a sea surface region and All India
Monsoon Rainfall. Also find lower and upper bounds of correlation coefficients at 99%
significant level. Give inferences. (Plot wherever necessary)

Plot correlation values over each grid of India with the temperature over sea surface.

Variables: Indian monsoon for 100 years (provided), Sea surface temperature for 100 years
(SST) (provided)

Group 4: Extract 100 years time series for sea surface temperatures over following regions:

(Everything right to Y-axis is East)

SST Region Co-ordinates of SST region


Bay of Bengal 6°N- 20°N, 80°E - 100°E

Niño3 5° S - 5° N, 210° E- 270° E

Niño4 5° S-5° N, 160° E - 210° E

North Atlantic 30° N - 45°N, 300° E - 350° E

Find the trend for 100 years. Find the trend for first 50 and last 50 years separately. Check if
the trends are significant or not. Infer from your analysis. Show all the plots.

(Remember, the co-ordinates given above will not exactly match with the data given. You
have to convert according to grid size of the data provided)

Variables: Sea surface temperature for 100 years (provided)

Group 5: Extract 100 years time series for sea surface temperatures over following regions:

(Everything right to Y-axis is East)

SST Region Co-ordinates of SST region

Niño3.4 5° S -5° N, 190° E- 240°E

South China Sea 5° N - 15°N, 110° E- 120°E

Southern Indian Ocean 5° S - 15° N, 40° E- 100°E

Western Indian Ocean 5° S-10° N,50° E- 65° E

Fit a linear regression model with sea surface regions, as input variables (predictors) and
rainfall as output variable. Show the model output and explain all the statistical parameters
computed from the model. Check the multicollinearity between all the sea surface dataset
(predictors). What changes should be made in the model, if necessary? (Plot wherever
necessary)

(Remember, the co-ordinates given above will not exactly match with the data given. You
have to convert according to grid size of the data provided)

Group 6: Fit a regression model (try multiple) over the given dataset in excel sheet. Find the
relationship between Net irrigated area (output) and other variables (input) with their
interactions. Specify which type of model fits the best and why. State if there is
heteroscedasticity in the input variables. (Plot wherever necessary)
Group 7: Extract the rainfall data specified for Group 1. Collect the rainfall dataset for
months June, July, August and September for all the years. Find whether the rainfall dataset
of each month belongs to same population or not. (Plot wherever necessary)

Group 8: A reservoir has some inflow Qi in m3/s. It has 9 gates for which outflow is given
Qo m3/s. Each observation is for a time step where a gate opens and discharges water. Fit an
interaction model with level of water as output variable and input as rest of the variables.
Explain the model and infer. (Plot wherever necessary)

Variables: Water level, Qi, Qo, Gate numbers (provided in same order as mentioned here)

Group 9:The discharge in a canal is Q. Width, depth and velocity variables are given. Find
multicollinearity within the predictors. Fit an interactive regression model with input
variables and predictands. Find on how many variables Q is dependent on. (Plot wherever
necessary)

Variables: Q, width, depth and velocity (provided in same order as mentioned here)

Group 10:A groundwater well storativity is dependent on the discharge of water from an
aquifer into the well (aquifer transmissivity), Radius of Influence and pumping rate. Fit
multiple models (e.g. linear, quadratic, cubic with their interactions) on the dataset provided.
Find if there is any multicollinearity within the input variables. (Plot wherever necessary)

Variables: Well storativity, Q, radius of influence and pumping rate (provided in same order
as mentioned here)

Procedure:

How to download NCEP-NCAR data? Find the variable name suggested in the question.
There will be several time related datasets such as daily or monthly. You can download any
of them but make sure to convert the data as per the question requirements.

How to read netcdf (.nc) files:

The files will be in the format of ‘.nc’ or ‘.nc4’ format. To display the content in these files in
MATLAB, you can use ncdisp() command.

Example: If your file name is ‘sea_surface_temperature.nc’, use


ncdisp(‘sea_surface_temperature.nc’) command to display the contents in the file. You will
find short names of variables which can be used to extract data. E.g. ‘lat’ for latitude, ‘lon’
for longitude or ‘pr’ for precipitation etc.

To extract the data into a .mat file, use ncread() function.

Example: If the variable you want to extract from ‘.nc’ file is ‘sst’ from filename
‘sea_surface_temperature.nc’, use aa = ncread(‘sea_surface_temperature.nc’,’sst’) where ‘aa’
will be your .mat filename.

You might also like