Pygeohydro

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

Bug Fixes
~~~~~~~~~
- In ``nlcd_helper`` function the roughness value for class 82 was set to 0.16
instead of 0.037.

New Features
~~~~~~~~~~~~
- Converted all methods of ``NWIS`` class to ``classmethod`` so the class can be used
without instantiating it. This change makes the class more flexible and easier to use.
- In ``NID`` class, the ``stage_nid_inventory`` method now checks if the remote NID
database has been modified since the last download and only downloads the new data
if it has been modified. This change makes the method more efficient and reduces the
network traffic while ensuring that the local database is always up-to-date.

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

Breaking Changes
~~~~~~~~~~~~~~~~
- Bump the minimum supported version of ``shapely`` to 2.

Internal Changes
~~~~~~~~~~~~~~~~
- Update the link to NWIS error codes tables in the ``nwis_errors`` function.
- Update ``NWIS`` class based on the latest changes to the NWIS web service.
- Use the default tiles for the ``interactive_map`` function.

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

New Features
~~~~~~~~~~~~
- Add a new attribute to ``EHydro`` class called ``survey_grid``.
It's a ``geopandas.GeoDataFrame`` that includes the survey grid
of the eHydro dataset which is a 35-km hexagonal grid.
- Add support for getting point cloud and survey outline data from
eHydro. You can set ``data_type`` in ``EHydro`` to ``bathymetry``,
``points``, ``outlines``, or ``contours`` to get the corresponding
data. The default is ``points`` since this is the recommended data
type by USACE.
- Add ``NFHL`` class within ``nfhl`` module to access FEMA's
National Flood Hazard Layer (NFHL) using six different ArcGISRESTFul
services. Contributed by
`Fernando Aristizabal <https://github.com/fernando-aristizabal>`__.
(:pull_hydro:`108`)

Internal Changes
~~~~~~~~~~~~~~~~
- Remove dependency on ``dask``.
- Move all NLCD related functions to a separate module called ``nlcd``.
This doesn't affect the API since the functions are still available
under ``pygeohydro`` namespace.

-------------------
This release provides access to three new datasets:

- USACE Hydrographic Surveys (eHydro) and
- USGS Short-Term Network (STN) Flood Event Data,
contributed by `Fernando Aristizabal <https://github.com/fernando-aristizabal>`__.
(:pull_hydro:`108`)
- NLCD 2021

New Features
~~~~~~~~~~~~
- Add support for getting topobathymetry data from USACE Hydrographic
Surveys (eHydro). The new class is called ``EHydro`` and gives users
the ability to subset the eHydro dataset by geometry, ID, or SQL queries.
- Add new ``stnfloodevents`` module with ``STNFloodEventData`` class for
retrieving flood event data from the
`USGS Short-Term Network (STN) <https://stn.wim.usgs.gov/STNWeb/#/>`__
RESTful Service. This Python API abstracts away RESTful principles and
produces analysis ready data in geo-referenced GeoDataFrames, DataFrames,
lists, or dictionaries as desired. The core class methods available are
``data_dictionary``, ``get_all_data``, and ``get_filtered_data``.
These class methods retrieve the data dictionaries by type, get all the
available data by type, and make filtered requests for data by type as well,
respectively. The four types of data include ``instruments``, ``peaks``,
``hwms``, and ``sites``.
Contributed by `Fernando Aristizabal <https://github.com/fernando-aristizabal>`__.
- Add a wrapper function for the ``STNFloodEventData`` class called
``stn_flood_event``.
- Add support for the new NLCD data (2021) for the three supported layers.

-------------------
From release 0.15 onward, all minor versions of HyRiver packages
will be pinned. This ensures that previous minor versions of HyRiver
packages cannot be installed with later minor releases. For example,
if you have ``py3dep==0.14.x`` installed, you cannot install
``pydaymet==0.15.x``. This is to ensure that the API is
consistent across all minor versions.

New Features
~~~~~~~~~~~~
- Add a new option to ``NWIS.get_info``, called ``nhd_info``, for
retrieving NHDPlus related info on the sites. This will two new
service calls that might slow down the function, so it's disabled
by default.
- Update links in ``NID`` to the latest CSV and GPKG versions of
the NID dataset.
- Add two new properties to ``NID`` to access the entire NID dataset.
You can use ``NID.df`` to access the CSV version as a
``pandas.DataFrame`` and ``NID.gdf`` to access the GPKG version
as a ``geopandas.GeoDataFrame``. Installing ``pyogrio`` is highly
recommended for much faster reading of the GPKG version.
- Refactor ``NID.bygeom`` to use the new ``NID.gdf`` property for
spatial querying of the dataset. This change should make the query
much faster.
- For now, retain compatibility with ``shapely<2`` while supporting
``shapley>=2``.

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

New Features
~~~~~~~~~~~~
- Add a new function, called ``nlcd_area_percent``, for computing the
percentages or natural, developed, and impervious areas within geometries
of a given ``GeoDataFrame``. This function uses imperviousness and land
use/land cover data from NLCD to compute the area percentages of the natural,
developed, and impervious areas. For more information please refer to the
function's documentation.
- Add a new column to the dataframe returned by ``NWIS.get_info``, called
``nhd_comid``, and rename ``drain_sqkm`` to ``nhd_areasqkm``. The new
drainage area is the best available estimates of stations' drainage area
that have been extracted from the NHDPlus. The new ``nhd_comid`` column
makes it easier to link stations to NHDPlus.
- In ``get_camels``, return ``qobs`` with negatives values set to ``NaN``.
Also, Add a new variable called ``Newman_2017`` to both datasets for
identifying the 531 stations that were used in
`Newman et al. (2017) <https://doi.org/10.1175/JHM-D-16-0284.1>`__.
- Add a new function, called ``streamflow_fillna``, for filling missing
streamflow values (``NAN``) with day-of-year average values.

Breaking Changes
~~~~~~~~~~~~~~~~
- Bump the minimum required version of ``shapely`` to 2.0,
and use its new API.

Internal Changes
~~~~~~~~~~~~~~~~
- Sync all minor versions of HyRiver packages to 0.14.0.
- Improve performance of all NLCD functions by merging two methods of
the ``NLCD`` and also reducing the memory footprint of the functions.