README

This contains all the software used by the current version of est_noise.
est_noise is a linear regression program to estimates various
functional relations in time seriers (ie, rate, offsets, etc) and,
importantly, estimates the parameters that describe the temporal correlations
such that the estimated values of the functional relations are 'relatively'
unbiased (ie, representing the true values).

This package includes "the manual".
There is a compile.sh script which allows
one to use either gfortran or Intel fortran to create the executables
which will land in the bin directory.

The directory example provides the input file and sample data to test est_noise.
The input file is anotated .. please read the manual

The directory example0 provides some scripts that use all of the
executables to assist with analyzing time series data. Please read the manual.

For the interested user, there are three other reports provided:
Langbein_Geod_2017.pdf  -- explains the method of how est_noise works
Threshold_dMLE_v4.pdf  -- describes how to select the "best" noise model
     that models the temporal correlation
GGMproblemFixed.pdf  -- describes the modification of the original est_noise (v7.22)
    to fix the problem of incorrectly computing the rate and its uncertainty when
    a Gauss-Markov process is used to describe the temporal correlations.

Older versions of est_noise can be found at:
https://escweb.wr.usgs.gov/share/langbein/Web/OUT/est_noise/
and
https://escweb.wr.usgs.gov/share/langbein/Web/OUT/est_noise_legacy/

Questions --- contact me;
langbein@usgs.gov
john_langbein@yahoo.com

March 31, 2023