<< Click to Display Table of Contents >> Navigation:  »No topics above this level« Interferometric Stacking - Stacking Tools - Tomographic Slice

Purpose

 

This tool uses the SBAS raster deformation time series to generate a classification map by displacement types. It tries different curve fittings selecting the most appropriate one using the Akaike information criterion.

 

Two types of fitting models are available: analytical or phenomenological. The available analytical fittings are linear, quadratic or periodical. The phenomenological models rely on external data provided by the user such as rain, temperature, water levels or any other quantity relevant to the study. More than one may be provided at the same time.

 

The output is a classification map, the best fitting parameters for each model and an optional fitted time series per model.

 

Technical Notes

 

Input Files

 

Input File

The displacement meta file. Both decomposed or LOS meta files are supported

 

Optional Files

Input Secondary File List

A list of external measurements in the form of csv and/or meta files such as the ones obtained from the Era5 or ECMWF download tools.

 

The csv files may include a header which is skipped while reading and the data lines must be in the following format:

 

d-m-yyyy,x

 

with

 

 d the day number

 m the month number

 yyyy the year

 - the date separator

 , the column separator

 x the measured quantity

 

Example:

 

date,rain

1-1-2014,86.2

1-2-2014,73.3

1-3-2014,51.8

1-4-2014,15.5

…

 

Parameters - Principal Parameters

 

Generate fitted TS

By setting this flag the fitted time series for each model will be saved. This may be useful to analyze different model performance.

 

Linear fit

By setting this flag the linear model fitting will be performed.

 

Quadratic fit

By setting this flag the quadratic model fitting will be performed.

 

Periodical fit

By setting this flag the periodical model fitting will be performed.

 

External parameter fit

By setting this flag the external parameter (phenomenological) model fitting will be performed.

 

Classify

By setting this flag a classification file according to the best fitting models will be saved.

 

L1 threshold

The L1 norm displacement threshold to consider a pixel as a stable one. Pixels under this threshold will be considered as no displacement points in the classification map.

 

RMSE threshold

The maximum root mean square of the fitting allowed to consider a valid fitting. Pixels over this RMES will be displayed ad not classified on the classification map.

 

Other parameters

In the other parameters section limits for the possible fitting parameters can be introduced. If a parameter if forced fixed by setting the maximum value equal to the minimum value, the model will be considered as a lower order model in the Akaike classification.  

Output Files

 

All out files, except for the time series which are in their corresponding directory, are named using the Output Root Name. These files are:

 

rootName_classification

Displacement classification file.

 

rootName_modelName_aN

Image with fitting parameter of order N

 

rootName_modelName_chi

Image with the fitting chi

 

rootName_modelName_rms

Image with the fitting rms error

 

rootName_modelName_TS (directory)

 

 modelName_meta

 Meta file of the fitted time serire

 

 modelName_input_TS

 Fitted time series images

General Functions

 

Cancel        

The window will be closed.

 

Help        

Specific help document section.

 

 

 

Specific Function(s)

 

None.

 

References