FIMsim/environment.yml at main · sdmlua/FIMsim · GitHub

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
name: FIMsim
channels:
  - conda-forge
  - defaults
dependencies:
  - python=3.11

  # Geospatial stack via conda-forge — all share the same GDAL, no DLL conflicts
  - rasterio
  - pyproj
  - pyogrio
  - shapely
  - fiona
  - geopandas
  - rioxarray

  # Scientific / numerical
  - numpy
  - scipy
  - pandas
  - openpyxl
  - matplotlib-base   # NOT matplotlib — avoids pulling in Qt/PySide6 from conda

  # NetCDF / HDF / cloud storage (NWM retrospective)
  - xarray
  - zarr
  # s3fs/aiobotocore must stay compatible with the botocore that boto3 and
  # awscli pin, or anonymous S3 breaks and FIMserv (OWP HAND-FIM) silently
  # downloads no NWM discharge — see the pip section below.
  - s3fs
  - fsspec
  - numcodecs
  - netcdf4
  - h5netcdf
  - h5py
  - aiofiles
  - aiobotocore

  # Hydrology data download (NHD flowlines, LULC, NWM)
  - pynhd
  - pygeoogc
  - pygeoutils
  - async-retriever

  # ARC-Curve2Flood backend — NenCarta drives ARC + Curve2Flood.
  # gdal gives the osgeo Python bindings these tools import; conda-forge keeps it
  # matched to the same GDAL as rasterio (no C-library conflict).
  - gdal
  - numba=0.60.0   # pinned: fimserve requires exactly this
  - fastparquet
  - geojson
  - cython
  - whitebox

  # HTTP / networking
  - requests
  - certifi
  - urllib3
  - aiohttp

  - pip
  - pip:
      - pyqt6==6.7.1   # pip-only — conda's Qt6 would conflict with this
      - fimserve        # OWP HAND-FIM backend (HAND-FIM mode)
      - "zarr>=3.0.1"  # fimserve's kerchunk dep requires zarr 3.x; pin to prevent legacy-resolver downgrade
      # ARC-Curve2Flood — GitHub-only (not on PyPI). --no-deps not needed here
      # because the conda block above already provides gdal/numba/etc.
      - "git+https://github.com/MikeFHS/automated-rating-curve.git"   # ARC (import arc)
      - "git+https://github.com/MikeFHS/Curve2Flood.git"             # Curve2Flood (import curve2flood)
      # NenCarta orchestrates ARC + Curve2Flood and provides the `flood-mapping`
      # CLI that FIMsim's ARC-Curve2Flood mode drives (see its README).
      - geoglows
      - progress
      # NenCarta imports PyQt5 at module load (nencarta/gui_app.py).  FIMsim
      # itself is PyQt6 and never imports nencarta in-process — it shells out to
      # `flood-mapping` — and main.py pins QT_API=pyqt6 so matplotlib cannot
      # bind to PyQt5.
      - "PyQt5>=5.15"
      - "git+https://github.com/jlgutenson/nencarta.git"             # flood-mapping CLI
      # ── Anonymous-S3 stack (FIMserv reads the NWM retrospective zarr from
      # S3 via s3fs).  aiobotocore pins botocore VERY tightly, and boto3/awscli
      # pin it too; when the two pins disagree every anonymous S3 read dies with
      #   TypeError: ClientArgsCreator.compute_endpoint_resolver_builtin_defaults()
      #   missing 1 required positional argument: 's3_disable_express_session_auth'
      # and FIMserv then generates FIM with no discharge — i.e. no flood map.
      # aiobotocore 3.9 accepts botocore 1.43.3–1.43.56 (boto3 1.43.36 pins
      # 1.43.36), and s3fs >= 2026.1 allows aiobotocore 3.x.  Keep these three
      # consistent if you ever upgrade one.
      - "aiobotocore>=3.9.0"
      - "s3fs>=2026.2.0"