release_note.txt

=====================
isce-2.3
=====================

First version of the open source release of ISCE under the Apache License-2.0.
Please read the LICENSE file included in this package.

ISCE is the InSAR Scientific Computing Environment.  It provides a framework
for constructing components and applications that implement workflows, in
general, with broad built-in support for processing InSAR data from raw inputs
through to geocoded interferograms with options for filtering, unwrapping, and
many more standard products.  The InSAR workflows are located in the applications
directory.  Example input files for the applications are given in the directory,
examples/input_files.  Some examples for constructing and using applications are
given in the examples/applications directory.

ISCE is developed with Python3 components and workflows that access legacy and
newer libraries that are developed in Fortran and C/C++.  Please read the README
file for information on the software dependencies that must be installed and
environment settings that must be set in order to use ISCE.


=====================
isce-2.2.1
=====================

Planned final release before open source release.

The next release of isce-2 is planned be an open source release.  Details on where to access
it will be coming soon. Code elements required for the insarApp.py and isceApp.py workflows
will continue to require a license in order to use them.  The open source version of ISCE
will not include those code elements.  Information will be made available for those people
wanting access to the Stanford licensed components.  The open source version includes
workflows such as stripmapApp.py that are included in this release.

Additions and Bug Fixes in this release:

* Fixed bug in mdx wrap interpretation - contributed by Gerald Manipon
* Fixed bug in isceobj.Image with ENVI Hdr Render - contributed by Gerald Manipon

* applications/downsampleDEM.py: contributed by David Bekaert
    - enforcing grid range to be a multiple of the sampling
    - adding option to either specify resolution in meters or in arcsecond
      Arc-seconds ensure that one can sample degrees consistently.
    - topsApp.py changes: tight burst validity required only when run with ionopsheric estimation on.
      This remove the issue of gaps between two separate along-track processed topsApp products.

* topsApp: updated to include ionospheric corrections - contributed by Cunren Liang
   - modified: components/isceobj/Sensor/TOPS/BurstSLC.py
               components/isceobj/Sensor/TOPS/Sentinel1.py

* contrib/alos2proc: contributed by Cunren Liang

* GPUTopozero: changed byte offsets to long type

* applications/gdal2isce_xml.py: removed the hard coded scheme of the data in favor of directly capturing it from gdal

* topsApp.xml: minor updates (see examples/input_files/reference_TOPS_SENTINEL1.xml)


=====================
isce-2.2.0
=====================

-- General bug fixes
-- Support for GDAL + c++11 starting from GDAL2.3
-- Added a --skipcheck option which will disable raw_input if optional dependencies like HDF5
   are not found. This was a feature requested by folks dockerizing ISCE for use on cloud
   machines.
-- Include option to specify VRT in gdal2isce_xml.py
-- Added conversion scripts from stack output to stamps input
-- Fixed renderVRT to not include geotransform or SRS when rendering for non-geocoded images.
   Currently only mechanism to distinguish radar / geocoded images is by looking at coord1
   and coord2 members of Image class
-- For Sentinel1, increased orbit extraction from 40 to 60 seconds
-- Changed roiApp.py/roiProc to stripmapApp.py/stripmapProc
   Renamed examples/input_files/roiApp.xml to stripmapApp.xml
-- Added more checks when reading in new format SLCs
   No guarantees that the reader works for all new variants
-- For topsApp added option to use a different DEM for geocoding.
   If none is provided, defaults to the DEM used for processing topo.

------------------------------

* Updates related to topsApp:
  -- Bug fixes corresponding to valid sample computation in range
  -- Warning messages generated for data acquired during orbit maneuvers
  -- Option of using a different dem for geocoding via the property “geocode demfilename”.
     Suitable for data with large number of looks. User can use downsampleDEM.py to generate
     an appropriately downlooked DEM for use with geocoding.

* Bug fix to VRT files. EPSG information is now not included for radar geometry files.

* Updated version of pyCuAmpcor from Caltech with downstream search capability

* Stripmap readers
  -- Sentinel-1 stripmap reader can read from zip files just like the TOPS reader
  -- Updates to Envisat SLC reader to support new format from ESA
  -- Updates to ICU unwrapper to output multiple connected components (used to be the one
     largest component earlier)
  -- Scons can now either be python2 / python3 based.
  -- ALOS2 ScanSAR reader (use VRTs and save disk space without replicating data). Not used
     by any app currently.

---------------------------------

stripmapApp.py:

* The roiApp.py in the previous release, has been renamed to “stripmapApp.py” for clarity.
As the self-descriptive “stripmapApp.py” implies, this ISCE application can be used for
interferometric processing of stripmap SAR data acquired by a wide range of sensors.
Several bugs have been fixed compared to the previous release and the stripmapApp has been
tested on different datasets (including ALOS-1 and ALOS-2 stripmap data, CSK, Sentinel-1
stripmap and Envisat data).

* The ionospheric phase estimation have been significantly improved by enforcing common
phase unwrapping error in the sub-band interferograms.

*** There are additional README files in the contrib/stack directory

---------------------------------
Stack processing with ISCE:

We have added standalone scripts to contrib/stack folder, which allows advanced users to
process a stack of SAR data using ISCE. These scripts are experimental and have been mainly
developed for prototyping different algorithms (e.g., precise coregistration and ionospheric
phase estimation). However, due to large interest of the InSAR community, we have decided
to release this scripts as is. These scripts won’t be automatically included in your install
directory after installing ISCE. However, you can easily use them based on the instructions
found in  contrib/stack/INSTALL.txt in your source ISCE directory.

Note that there are two different stack processors for processing a stack of TOPS or stripmap
data, found in  “contrib/stack/topsStack” and “contrib/stack/stripmapStack” respectively.


In the “contrib/stack/“ folder there are three text files:

INSTALL.txt:               explains how users should setup their $PATH variable to use the
                           stack processor
README_topsStack.txt:      instructions for using the scripts in topsStack directory for
                           processing a stack of Sentinel-1 TOPS data
README_stripmapStack.txt:  instructions for using the scripts in stripmapStack directory for
                           processing a stack of stripmap data

---------------------------------

=====================
isce-2.1.0 (20170803)
=====================

NEW FEATURES. HIGHLIGHTS:

-- Added ERS_EnviSAT_SLC reader. Contributed by Scott Henderson from the ISCE
user community.  Thank you, Scott!

-- Change to minimum gcc requirement.  Must be >= gcc4.7.  Compilation of
C/C++ code is using the "-std=c++11", which requires >= gcc4.7.  See the README
file for required versions of software dependencies."

-- New RoiProc and roiApp.py workflow added.  New code to estimate ionospheric
contributions to phase is included in contrib/splitSpectrum and used in roiApp.py.

-- The "steps" mode of processing has a new command line option: "--next".  This
option determines the next step for a given application based on the most
recently updated file in the PICKLE folder, whether the files are pickle files
or xml files; the setting for the property "renderer" determines which type of
step file is used for determining the next step to be processed using --next
(the default being pickle files).

-- The "steps" mode now renders the xxxProc.xml (as in insarProc.xml) file after
each step so that it is available to view/use after each step.  It will be
overwritten after each step completes.

-- Added a new simplified water mask downloader, "wbd.py", and dem downloader,
"demdb.py". The only required command line arguments are the "snwe" lat, lon
coordinates. They create stitched watermask and dem files with xml and vrt
meta data files.  If you use the $DEMDB environment variable the files will
be moved to the directory pointed to by $DEMDB and will be automatically
discoverable by ISCE anywhere on your system.

-- topsApp.py DEM downloading now downloads to the directory pointed to by your
$DEMDB environment variable (like insarApp.py and isceApp.py do).  If $DEMDB is
not defined, then the files will be downloaded in the local processing directory.
The dems and water masks in the $DEMDB directory are available for use in any
directory automatically.  If you want to move your prviously downloaded dems and
water masks to the $DEMDB directory you will have to edit the paths in their xml
files.

-- This release includes some code elements that can run on a GPU requiring as
dependencies the CUDA API and also an executable named 'cython3' to invoke
cython.  You may need to manually create a soft link with this name pointing
to the name given to your installed cython.  The code that require these new
dependencies are in components/zerodop/GPUtopozero and GPUgeo2rdr directories.
This code is not compiled and the CUDA and cython3 dependencies are not
required unless you include the following line in your SConfigISCE file:

ENABLE_CUDA=True

If you do not include this line or if you set the value to False, then
the code will compile as usual.  This is experimental code at this time
and is not used in the builtin Applications.  More code using CUDA for
GPUs will be coming in the future.

-- Added 'useGPU' configuration parameter to topsApp.py to use the GPU code
mentioned above in "New Features".  You must also include the "ENABLE_CUDA=True"
line in your SConfigISCE file in order to compile the CUDA code.

-- Added "extra ESD cycles" parameter to topsApp.py to add extra ESD cycles
to interpret overlap phase.



GENERAL BUG FIXES.  HIGHLIGHTS:

-- replaced dependency on "cv2" with "scipy" in RoiProc/runDispersive.py

-- fixed open file leaks in image file accessors and added more output to
show methods being used to open and close files.


=============================================================================
Reminder of previous new capabilities and bug fixes:
1. topsApp related
    a. Can provide zip files / safe folders as inputs
    b. Bug fixes to support multi slice / region of interest processing
    c. Generation of traditional coherence products
    d. Output file names match insarApp – single resolution merged products
       have .full suffix
    e. Looks, filter strength, thresholds etc moved from topsproc level to
       topsapp level
    f. Changes to topsApp to use new Image API decorators
    g. Added support for Sentinel-1B
2. topsOffsetApp
    a. To be run in same folder as topsApp.
    b. Produces dense pixel offset maps from Sentinel data.
3. Stripmap processing related
    a. Stripped ALOS raw data of headers
    b. Fixed bug in stripmap frame stitcher when starting ranges were different
4. General changes
    a. Bug fixes to support near-nadir imaging geometries like SWOT
    b. Geocoding support for datatypes other than FLOAT / CFLOAT
    c. Updates to envisat reader to handle new ESA format
    d. Bug fix in water mask generation when input DEM doesn’t span the image.
    e. Bug fix in DemStitchers to enable dem.py to work again.

===============================================================================
Previous note about compiling ISCE after unsuccessful build due to missing headers
and libraries:

When building ISCE, scons will check the list of header files and libraries that
ISCE requires.  Scons will cache the results of this dependency checking.  So,
if you try to build ISCE and scons tells you that you are missing headers or
libraries, then you should remove the cached files before trying to build ISCE
again after installing the missing headers and libraries.  The cached files are
config.log, .sconfig.dblite, and the files in directory .sconf_temp.  You should
run the following command while in the top directory of the ISCE source (the
directory containing the SConstruct file):

> rm -rf config.log .sconfig.dblite .sconf_temp

and then try "scons install" again.

===============================================================================
From past notes about automatic downloading of dems:

1. You need to have a user name and password from urs.earthdata.nasa.gov and you
need to include LPDAAC applications to your account.
    a. If you don't already have an earthdata username and password,
       you can set them at https://urs.earthdata.nasa.gov/
    b. If you already have an earthdata account, please ensure that
       you add LPDAAC applications to your account:
         - Login to earthdata here: https://urs.earthdata.nasa.gov/home
         - Click on my applications on the profile
         - Click on “Add More Applications”
         - Search for “LP DAAC”
         - Select “LP DAAC Data Pool” and “LP DAAC OpenDAP” and approve.

2. create a file named .netrc with the following 3 lines:

machine urs.earthdata.nasa.gov
    login your_earthdata_login_name
    password your_earthdata_password


3. set permissions to prevent others from viewing your credentials:

> chmod go-rwx .netrc

===============================================================================
From past note on making stitched dems and water masks globally available:

Stitched dems and water body masks will be stored in and used from a directory
indicated by the environment variable, $DEMDB.  If you define this environment
variable with value equal to a path where you want to store stitched dems and
waterbody masks, then any stitched dem or water mask will be globally available
automatically without needing to specify any information about the dem in your
input files for ISCE processing applications. If you use dem.py or watermask.py,
the stitched products are left in the directory where you run these apps.  If
you want them to be globally available, then either run dem.py or watermask.py
in the $DEMDB directory or else move them there.

==============================================================================
From past note on where to read about installing and running ISCE and on how
to obtain auxiliary files needed for orbit computations when working with RadarSAT1
data.

The README.txt file and the example input files in the examples/input_files
directory are the most up to date sources of information on installing and running
ISCE. The ISCE.pdf file is useful but may be dated in parts.


If you want to use ISCE to process RadarSAT1 data, then you will need to download
files to the components/isceobj/Orbit/db directory. Instructions on where to get
the files are given in the file kernels.list file in that directory.

===================
End-Of-Release-Note
===================