diff --git a/.github/workflows/tests.yml b/.github/workflows/ci.yml
similarity index 53%
rename from .github/workflows/tests.yml
rename to .github/workflows/ci.yml
index 1e0de81ce2..cbedcb778f 100644
--- a/.github/workflows/tests.yml
+++ b/.github/workflows/ci.yml
@@ -1,13 +1,16 @@
-name: Tests
+name: CI Tests
 
-on: [pull_request, push]
+on:
+  pull_request:
+  push:
+  workflow_dispatch:
 jobs:
   Sphinx-Pytest-Coverage:
     runs-on: ubuntu-latest
     strategy:
       fail-fast: false
       matrix:
-        env: [py37_iris30, py38_iris30]
+        env: [environment_a, environment_b, conda-forge]
     steps:
     - uses: actions/checkout@v2
     - uses: actions/cache@v2
@@ -16,45 +19,54 @@ jobs:
         # Increase this value to reset cache
         CACHE_NUMBER: 2
       with:
-        path: /usr/share/miniconda/envs/improver_${{ matrix.env }}
-        key: ${{ format('{0}-conda-improver-{1}-{2}-{3}', runner.os, env.CACHE_NUMBER, matrix.env, hashFiles(format('envs/environment_{0}.yml', matrix.env))) }}
+        path: /usr/share/miniconda/envs/im${{ matrix.env }}
+        key: ${{ format('{0}-conda-improver-{1}-{2}-{3}', runner.os, env.CACHE_NUMBER, matrix.env, hashFiles(format('envs/{0}.yml', matrix.env))) }}
     - name: conda env update
       if: steps.cache.outputs.cache-hit != 'true'
       run: |
         source '/usr/share/miniconda/etc/profile.d/conda.sh'
         conda install -c conda-forge mamba
-        mamba env update -q --file envs/environment_${{ matrix.env }}.yml --name improver_${{ matrix.env }}
+        mamba env update -q --file envs/${{ matrix.env }}.yml --name im${{ matrix.env }}
     - name: conda info
       run: |
         source '/usr/share/miniconda/etc/profile.d/conda.sh'
-        conda activate improver_${{ matrix.env }}
+        conda activate im${{ matrix.env }}
         conda info
         conda list
-    - name: sphinx-build
+    - name: sphinx documentation
       run: |
         source '/usr/share/miniconda/etc/profile.d/conda.sh'
-        conda activate improver_${{ matrix.env }}
+        conda activate im${{ matrix.env }}
         make -C doc html SPHINXOPTS="-W --keep-going"
-    - name: pytest unit-tests & cov-report
+    - name: pytest without coverage
+      if: matrix.env != 'environment_a'
       run: |
         source '/usr/share/miniconda/etc/profile.d/conda.sh'
-        conda activate improver_${{ matrix.env }}
+        conda activate im${{ matrix.env }}
+        pytest
+    - name: pytest with coverage
+      if: matrix.env == 'environment_a'
+      run: |
+        source '/usr/share/miniconda/etc/profile.d/conda.sh'
+        conda activate im${{ matrix.env }}
         pytest --cov=improver --cov-report xml:coverage.xml
-    - name: codacy-coverage
-      if: env.CODACY_PROJECT_TOKEN
+    - name: codacy upload
+      if: env.CODACY_PROJECT_TOKEN && matrix.env == 'environment_a'
       run: |
         source '/usr/share/miniconda/etc/profile.d/conda.sh'
-        conda activate improver_${{ matrix.env }}
+        conda activate im${{ matrix.env }}
         python-codacy-coverage -v -r coverage.xml
       env:
         CODACY_PROJECT_TOKEN: ${{ secrets.CODACY_PROJECT_TOKEN }}
-    - uses: codecov/codecov-action@v1
+    - name: codecov upload
+      uses: codecov/codecov-action@v1
+      if: matrix.env == 'environment_a'
   Codestyle-and-flake8:
     runs-on: ubuntu-latest
     strategy:
       fail-fast: false
       matrix:
-        env: [py37_iris30]
+        env: [environment_a]
     steps:
     - uses: actions/checkout@v2
     - uses: actions/cache@v2
@@ -63,41 +75,41 @@ jobs:
         # Increase this value to reset cache
         CACHE_NUMBER: 2
       with:
-        path: /usr/share/miniconda/envs/improver_${{ matrix.env }}
-        key: ${{ format('{0}-conda-improver-{1}-{2}-{3}', runner.os, env.CACHE_NUMBER, matrix.env, hashFiles(format('envs/environment_{0}.yml', matrix.env))) }}
+        path: /usr/share/miniconda/envs/im${{ matrix.env }}
+        key: ${{ format('{0}-conda-improver-{1}-{2}-{3}', runner.os, env.CACHE_NUMBER, matrix.env, hashFiles(format('envs/{0}.yml', matrix.env))) }}
     - name: conda env update
       if: steps.cache.outputs.cache-hit != 'true'
       run: |
         source '/usr/share/miniconda/etc/profile.d/conda.sh'
         conda install -c conda-forge mamba
-        mamba env update -q --file envs/environment_${{ matrix.env }}.yml --name improver_${{ matrix.env }}
+        mamba env update -q --file envs/${{ matrix.env }}.yml --name im${{ matrix.env }}
     - name: conda info
       run: |
         source '/usr/share/miniconda/etc/profile.d/conda.sh'
-        conda activate improver_${{ matrix.env }}
+        conda activate im${{ matrix.env }}
         conda info
         conda list
     - name: isort
       run: |
         source '/usr/share/miniconda/etc/profile.d/conda.sh'
-        conda activate improver_${{ matrix.env }}
+        conda activate im${{ matrix.env }}
         isort --check-only .
     - name: black
       run: |
         source '/usr/share/miniconda/etc/profile.d/conda.sh'
-        conda activate improver_${{ matrix.env }}
+        conda activate im${{ matrix.env }}
         black --check .
     - name: flake8
       run: |
         source '/usr/share/miniconda/etc/profile.d/conda.sh'
-        conda activate improver_${{ matrix.env }}
+        conda activate im${{ matrix.env }}
         flake8 improver improver_tests
   Safety-Bandit:
     runs-on: ubuntu-latest
     strategy:
       fail-fast: false
       matrix:
-        env: [py37_iris30, py38_iris30]
+        env: [environment_a, environment_b, conda-forge]
     steps:
     - uses: actions/checkout@v2
     - uses: actions/cache@v2
@@ -106,60 +118,27 @@ jobs:
         # Increase this value to reset cache
         CACHE_NUMBER: 2
       with:
-        path: /usr/share/miniconda/envs/improver_${{ matrix.env }}
-        key: ${{ format('{0}-conda-improver-{1}-{2}-{3}', runner.os, env.CACHE_NUMBER, matrix.env, hashFiles(format('envs/environment_{0}.yml', matrix.env))) }}
+        path: /usr/share/miniconda/envs/im${{ matrix.env }}
+        key: ${{ format('{0}-conda-improver-{1}-{2}-{3}', runner.os, env.CACHE_NUMBER, matrix.env, hashFiles(format('envs/{0}.yml', matrix.env))) }}
     - name: conda env update
       if: steps.cache.outputs.cache-hit != 'true'
       run: |
         source '/usr/share/miniconda/etc/profile.d/conda.sh'
         conda install -c conda-forge mamba
-        mamba env update -q --file envs/environment_${{ matrix.env }}.yml --name improver_${{ matrix.env }}
+        mamba env update -q --file envs/${{ matrix.env }}.yml --name im${{ matrix.env }}
     - name: conda info
       run: |
         source '/usr/share/miniconda/etc/profile.d/conda.sh'
-        conda activate improver_${{ matrix.env }}
+        conda activate im${{ matrix.env }}
         conda info
         conda list
     - name: safety
       run: |
         source '/usr/share/miniconda/etc/profile.d/conda.sh'
-        conda activate improver_${{ matrix.env }}
+        conda activate im${{ matrix.env }}
         safety check || true
     - name: bandit
       run: |
         source '/usr/share/miniconda/etc/profile.d/conda.sh'
-        conda activate improver_${{ matrix.env }}
+        conda activate im${{ matrix.env }}
         bandit -r improver
-  Type-checking:
-    runs-on: ubuntu-latest
-    strategy:
-      fail-fast: false
-      matrix:
-        env: [py37_iris30]
-    steps:
-    - uses: actions/checkout@v2
-    - uses: actions/cache@v2
-      id: cache
-      env:
-        # Increase this value to reset cache
-        CACHE_NUMBER: 2
-      with:
-        path: /usr/share/miniconda/envs/improver_${{ matrix.env }}
-        key: ${{ format('{0}-conda-improver-{1}-{2}-{3}', runner.os, env.CACHE_NUMBER, matrix.env, hashFiles(format('envs/environment_{0}.yml', matrix.env))) }}
-    - name: conda env update
-      if: steps.cache.outputs.cache-hit != 'true'
-      run: |
-        source '/usr/share/miniconda/etc/profile.d/conda.sh'
-        conda install -c conda-forge mamba
-        mamba env update -q --file envs/environment_${{ matrix.env }}.yml --name improver_${{ matrix.env }}
-    - name: conda info
-      run: |
-        source '/usr/share/miniconda/etc/profile.d/conda.sh'
-        conda activate improver_${{ matrix.env }}
-        conda info
-        conda list
-    - name: mypy
-      run: |
-        source '/usr/share/miniconda/etc/profile.d/conda.sh'
-        conda activate improver_${{ matrix.env }}
-        mypy improver || true
diff --git a/.github/workflows/scheduled.yml b/.github/workflows/scheduled.yml
new file mode 100644
index 0000000000..dcd2c412ca
--- /dev/null
+++ b/.github/workflows/scheduled.yml
@@ -0,0 +1,137 @@
+name: Scheduled Tests
+
+on:
+  schedule:
+    - cron: '7 4 * * *'
+  workflow_dispatch:
+jobs:
+  Sphinx-Pytest-Coverage:
+    runs-on: ubuntu-latest
+    strategy:
+      fail-fast: false
+      matrix:
+        env: [latest]
+    if: github.repository_owner == 'metoppv'
+    steps:
+    - uses: actions/checkout@v2
+    - uses: actions/cache@v2
+      id: cache
+      env:
+        # Increase this value to reset cache
+        CACHE_NUMBER: 2
+      with:
+        path: /usr/share/miniconda/envs/im${{ matrix.env }}
+        key: ${{ format('{0}-conda-improver-{1}-{2}-{3}', runner.os, env.CACHE_NUMBER, matrix.env, hashFiles(format('envs/{0}.yml', matrix.env))) }}
+    - name: conda env update
+      if: steps.cache.outputs.cache-hit != 'true'
+      run: |
+        source '/usr/share/miniconda/etc/profile.d/conda.sh'
+        conda install -c conda-forge mamba
+        mamba env update -q --file envs/${{ matrix.env }}.yml --name im${{ matrix.env }}
+    - name: conda info
+      run: |
+        source '/usr/share/miniconda/etc/profile.d/conda.sh'
+        conda activate im${{ matrix.env }}
+        conda info
+        conda list
+    - name: sphinx documentation
+      run: |
+        source '/usr/share/miniconda/etc/profile.d/conda.sh'
+        conda activate im${{ matrix.env }}
+        make -C doc html SPHINXOPTS="-W --keep-going"
+    - name: pytest without coverage
+      if: matrix.env != 'environment_a'
+      run: |
+        source '/usr/share/miniconda/etc/profile.d/conda.sh'
+        conda activate im${{ matrix.env }}
+        pytest
+    - name: pytest with coverage
+      if: matrix.env == 'environment_a'
+      run: |
+        source '/usr/share/miniconda/etc/profile.d/conda.sh'
+        conda activate im${{ matrix.env }}
+        pytest --cov=improver --cov-report xml:coverage.xml
+    - name: codacy upload
+      if: env.CODACY_PROJECT_TOKEN && matrix.env == 'environment_a'
+      run: |
+        source '/usr/share/miniconda/etc/profile.d/conda.sh'
+        conda activate im${{ matrix.env }}
+        python-codacy-coverage -v -r coverage.xml
+      env:
+        CODACY_PROJECT_TOKEN: ${{ secrets.CODACY_PROJECT_TOKEN }}
+    - name: codecov upload
+      uses: codecov/codecov-action@v1
+      if: matrix.env == 'environment_a'
+  Safety-Bandit:
+    runs-on: ubuntu-latest
+    strategy:
+      fail-fast: false
+      matrix:
+        env: [latest]
+    if: github.repository_owner == 'metoppv'
+    steps:
+    - uses: actions/checkout@v2
+    - uses: actions/cache@v2
+      id: cache
+      env:
+        # Increase this value to reset cache
+        CACHE_NUMBER: 2
+      with:
+        path: /usr/share/miniconda/envs/im${{ matrix.env }}
+        key: ${{ format('{0}-conda-improver-{1}-{2}-{3}', runner.os, env.CACHE_NUMBER, matrix.env, hashFiles(format('envs/{0}.yml', matrix.env))) }}
+    - name: conda env update
+      if: steps.cache.outputs.cache-hit != 'true'
+      run: |
+        source '/usr/share/miniconda/etc/profile.d/conda.sh'
+        conda install -c conda-forge mamba
+        mamba env update -q --file envs/${{ matrix.env }}.yml --name im${{ matrix.env }}
+    - name: conda info
+      run: |
+        source '/usr/share/miniconda/etc/profile.d/conda.sh'
+        conda activate im${{ matrix.env }}
+        conda info
+        conda list
+    - name: safety
+      run: |
+        source '/usr/share/miniconda/etc/profile.d/conda.sh'
+        conda activate im${{ matrix.env }}
+        safety check || true
+    - name: bandit
+      run: |
+        source '/usr/share/miniconda/etc/profile.d/conda.sh'
+        conda activate im${{ matrix.env }}
+        bandit -r improver
+  Type-checking:
+    runs-on: ubuntu-latest
+    strategy:
+      fail-fast: false
+      matrix:
+        env: [latest]
+    if: github.repository_owner == 'metoppv'
+    steps:
+    - uses: actions/checkout@v2
+    - uses: actions/cache@v2
+      id: cache
+      env:
+        # Increase this value to reset cache
+        CACHE_NUMBER: 2
+      with:
+        path: /usr/share/miniconda/envs/im${{ matrix.env }}
+        key: ${{ format('{0}-conda-improver-{1}-{2}-{3}', runner.os, env.CACHE_NUMBER, matrix.env, hashFiles(format('envs/{0}.yml', matrix.env))) }}
+    - name: conda env update
+      if: steps.cache.outputs.cache-hit != 'true'
+      run: |
+        source '/usr/share/miniconda/etc/profile.d/conda.sh'
+        conda install -c conda-forge mamba
+        mamba env update -q --file envs/${{ matrix.env }}.yml --name im${{ matrix.env }}
+    - name: conda info
+      run: |
+        source '/usr/share/miniconda/etc/profile.d/conda.sh'
+        conda activate im${{ matrix.env }}
+        conda info
+        conda list
+    - name: mypy
+      run: |
+        source '/usr/share/miniconda/etc/profile.d/conda.sh'
+        conda activate im${{ matrix.env }}
+        mypy improver || true
diff --git a/doc/source/Dependencies.rst b/doc/source/Dependencies.rst
new file mode 100644
index 0000000000..9a5a63d4c6
--- /dev/null
+++ b/doc/source/Dependencies.rst
@@ -0,0 +1,162 @@
+Dependencies
+================
+
+.. contents:: Contents
+    :depth: 2
+
+
+IMPROVER builds on the functionality provided by a range of open source
+libraries.
+
+Some of these libraries are widely used throughout the IMPROVER code, so are
+considered as required for IMPROVER as a whole.
+
+Other libraries are only used in specific parts of IMPROVER.
+These libraries are optional for IMPROVER as a whole, but are required to use
+related parts of IMPROVER.
+Optional installation of these libraries allows for smaller sized installations
+of IMPROVER and reduces conda environment dependency solving difficulties.
+
+Required
+-----------------
+
+cartopy
+~~~~~~~~~~~~~~~~~
+Cartopy is used for grid projections and coordinate transformations.
+
+https://scitools.org.uk/cartopy/docs/stable/
+
+
+cftime
+~~~~~~~~~~~~~~~~~
+cftime provides functions for handling time in NetCDF files according to the
+Climate and Forecast (CF) conventions.
+
+https://unidata.github.io/cftime/
+
+
+cf-units
+~~~~~~~~~~~~~~~~~
+cf-units provides units conversion following the Climate and Forecast (CF)
+conventions.
+
+https://cf-units.readthedocs.io/en/stable/
+
+
+Clize
+~~~~~~~~~~~~~~~~~
+Clize automatically generates command line interfaces (CLI) from Python function
+signatures.
+
+https://clize.readthedocs.io/en/stable/
+
+
+Dask
+~~~~~~~~~~~~~~~~~
+Dask lazy-loaded arrays are used (often via Iris) to reduce the amount of data
+loaded into memory at once.
+
+https://dask.org/
+
+
+Iris
+~~~~~~~~~~~~~~~~~
+Iris cubes are used as a primary data structure throughout the IMPROVER code.
+
+https://scitools-iris.readthedocs.io/en/stable/
+
+
+NetCDF4
+~~~~~~~~~~~~~~~~~
+Python library for reading and writing NetCDF data files via Iris.
+
+https://unidata.github.io/netcdf4-python/
+
+
+Numpy
+~~~~~~~~~~~~~~~~~
+Multidimensional numerical array library, used as the basis for Iris cubes and
+dask arrays.
+
+https://numpy.org/doc/stable/
+
+
+Scipy
+~~~~~~~~~~~~~~~~~
+Scientific python library, used for a variety of statistical, image processing,
+interpolation and spatial functions.
+
+https://docs.scipy.org/doc/scipy/reference/
+
+
+Sigtools
+~~~~~~~~~~~~~~~~~
+Sigtools provides introspection tools for function signatures.
+Sigtools is required by clize, so this dependency is needed anyway
+despite minor usage in IMPROVER.
+
+https://sigtools.readthedocs.io/en/stable/
+
+
+Sphinx
+~~~~~~~~~~~~~~~~~
+Sphinx is a documentation library for Python. IMPROVER requires it at runtime
+to generate help strings from CLI function docstrings via clize.
+
+https://www.sphinx-doc.org/en/master/
+
+
+Optional dependencies
+---------------------
+
+python-stratify
+~~~~~~~~~~~~~~~~~~
+Vectorised (cython) interpolation, particularly for vertical levels of the
+atmosphere.
+
+https://github.com/SciTools/python-stratify
+
+Required for CLIs: ``interpolate-using-difference``, ``phase-change-level``
+
+statsmodels
+~~~~~~~~~~~~~~~~~~
+Estimation of statistical models, used for
+:doc:`EMOS <improver.calibration.ensemble_calibration>`.
+
+https://www.statsmodels.org/stable/
+
+Required for CLIs: ``estimate-emos-coefficients``
+
+numba
+~~~~~~~~~~~~~~~~~~
+JIT compiler for numerical Python code, used for better computational performance.
+
+https://numba.readthedocs.io/en/stable/
+
+Required for CLIs: ``generate-timezone-mask-ancillary``
+
+PySTEPS
+~~~~~~~~~~~~~~~~~~
+Probabilistic nowcasting of radar precipitation fields, used for nowcasting.
+
+https://pysteps.github.io/
+
+Required for CLIs: ``nowcast-accumulate``, ``nowcast-extrapolate``,
+``nowcast-optical-flow-from-winds``
+
+pytz
+~~~~~~~~~~~~~~~~~
+Timezone database for Python.
+
+https://pythonhosted.org/pytz/
+
+Required for CLIs: ``generate-timezone-mask-ancillary``
+
+timezonefinder
+~~~~~~~~~~~~~~~~~~
+Lookup of timezone using geographic coordinates, used to generate timezone
+grids.
+
+https://timezonefinder.readthedocs.io/en/stable/
+
+Required for CLIs: ``generate-timezone-mask-ancillary``
diff --git a/doc/source/Running-at-your-site.rst b/doc/source/Running-at-your-site.rst
index 91b8c6fd50..ac7cea27a1 100644
--- a/doc/source/Running-at-your-site.rst
+++ b/doc/source/Running-at-your-site.rst
@@ -22,13 +22,10 @@ or ``python setup.py install`` (``python setup.py develop``). Note that
 ``pip install`` will not work in an empty environment due to problems
 with installation of the dependency ``iris`` via pip.
 
-Required dependencies are listed in
-`environment_py37_iris30.yml <https://github.com/metoppv/improver/blob/master/envs/environment_py37_iris30.yml>`_,
-and
-`environment_py38_iris30.yml <https://github.com/metoppv/improver/blob/master/envs/environment_py38_iris30.yml>`_,
-depending on your version requirements. Both environment files are used
-to run the test suite on Github actions, so these conda environments
-file should stay up to date with any dependency changes.
+Example environments are included in the repository ``envs`` directory.
+These environment files are used to run the test suite on Github actions,
+so they should stay up to date with any dependency changes. See also
+documentation about :doc:`use of dependencies in IMPROVER <Dependencies>`.
 
 Alternatively, you can manually 'install' by downloading the code and
 putting the IMPROVER ``bin/`` directory in your PATH.
diff --git a/doc/source/index.rst b/doc/source/index.rst
index a09e7f34cc..85e24660ab 100644
--- a/doc/source/index.rst
+++ b/doc/source/index.rst
@@ -15,6 +15,7 @@
    Guidance-for-Reviewing-Code
    How-to-implement-a-command-line-utility
    Running-at-your-site
+   Dependencies
    Ticket-Creation-and-Definition-of-Ready
 
 .. toctree::
diff --git a/envs/conda-forge.yml b/envs/conda-forge.yml
new file mode 100644
index 0000000000..e1d8793fed
--- /dev/null
+++ b/envs/conda-forge.yml
@@ -0,0 +1,28 @@
+# This environment is intended to be used for the conda-forge improver-feedstock
+# dependencies in https://github.com/conda-forge/improver-feedstock
+# If this file is changed, recipe/meta.yaml in the improver-feedstock repository
+# should also be updated.
+# This environment should not include optional dependencies.
+# This environment should pin versions so that all unit tests pass.
+name: improver_conda_forge
+channels:
+  - conda-forge
+dependencies:
+  - python>=3.6
+  # Included in improver-feedstock requirements
+  - cartopy<0.20
+  - cftime<1.5
+  - cf-units=2.1.5
+  - clize
+  - dask
+  - iris>=3.0,<3.1
+  - netCDF4
+  - numpy<1.21
+  - scipy
+  - sigtools
+  - sphinx
+  # Additional libraries to run tests, not included in improver-feedstock
+  - bandit
+  - filelock
+  - pytest
+  - sphinx-autodoc-typehints
diff --git a/envs/environment_py37_iris30.yml b/envs/environment_a.yml
similarity index 51%
rename from envs/environment_py37_iris30.yml
rename to envs/environment_a.yml
index e6f0e0a5d4..fff6eaf147 100644
--- a/envs/environment_py37_iris30.yml
+++ b/envs/environment_a.yml
@@ -1,32 +1,29 @@
-# Use this file to create a conda environment using:
-# conda env create --file envs/environment_py37_iris30.yml
-
-name: improver_py37_iris30
+# This environment is a representative example for usage of IMPROVER
+name: improver_a
 channels:
   - conda-forge
 dependencies:
   - python=3.7
   # Required
   - cartopy=0.19
-  - cftime<1.5
+  - cftime=1.2.1
   - cf-units=2.1.5
-  - clize
-  - dask
-  - iris=3.0
-  - netCDF4
-  - numpy<1.21
-  - pandas
-  - python-dateutil
-  - python-stratify
+  - clize=4.1.1
+  - dask=2021.8.1
+  - iris=3.0.3
+  - netCDF4=1.4.1
+  - numpy=1.20.2
+  - pandas=1.2.4
   - pytz=2020.5
-  - scipy=1.6
-  - sigtools
-  - sphinx
+  - scipy=1.6.2
+  - sigtools=2.0.2
+  - sphinx=4.0.1
   # Optional
-  - fastparquet
-  - numba
+  - fastparquet=0.7.1
+  - statsmodels=0.12.2
+  - numba=0.53.1
   - pysteps=1.4.1
-  - statsmodels
+  - python-stratify=0.1.1
   - timezonefinder=4.1.0
   # Development
   - astroid
@@ -35,7 +32,7 @@ dependencies:
   - codacy-coverage
   - filelock
   - flake8
-  - isort=5.*
+  - isort=5.7.0
   - mock
   - mypy
   - pytest
diff --git a/envs/environment_b.yml b/envs/environment_b.yml
new file mode 100644
index 0000000000..82057101bf
--- /dev/null
+++ b/envs/environment_b.yml
@@ -0,0 +1,41 @@
+# This environment is a representative example for usage of IMPROVER
+# It does not include dependencies for the following IMPROVER CLIs:
+# estimate-emos-coefficients, generate-timezone-mask-ancillary,
+# nowcast-accumulate, nowcast-extrapolate and nowcast-optical-flow-from-winds
+# See doc/source/Dependencies.rst for more information
+name: improver_b
+channels:
+  - conda-forge
+dependencies:
+  - python=3.8
+  # Required
+  - cartopy=0.19
+  - cftime=1.2.1
+  - cf-units=2.1.5
+  - clize=4.2.0
+  - dask=2021.8.1
+  - iris=3.0.3
+  - netCDF4=1.5.7
+  - numpy=1.20
+  - pandas=1.3.4
+  - pytz=2020.5
+  - scipy=1.6.2
+  - sigtools=2.0.3
+  - sphinx=4.0.2
+  # Optional
+  - numba=0.53.1
+  - python-stratify=0.2.post0
+  # Development
+  - astroid
+  - bandit
+  - black=19.10b0
+  - filelock
+  - flake8
+  - isort=5
+  - mock
+  - mypy
+  - pytest
+  - pytest-cov
+  - pytest-xdist
+  - safety
+  - sphinx-autodoc-typehints
diff --git a/envs/environment_py38_iris30.yml b/envs/environment_py38_iris30.yml
deleted file mode 100644
index 413f894951..0000000000
--- a/envs/environment_py38_iris30.yml
+++ /dev/null
@@ -1,40 +0,0 @@
-# Use this file to create a conda environment using:
-# conda env create --file envs/environment_py38_iris30.yml
-
-name: improver_py38_iris30
-channels:
-  - conda-forge
-dependencies:
-  - python=3.8
-  # Required
-  - cartopy=0.19
-  - cftime<1.5
-  - cf-units=2.1.5
-  - clize
-  - dask
-  - iris=3.0
-  - netCDF4
-  - numpy<1.21
-  - pandas
-  - python-dateutil
-  - python-stratify
-  - pytz=2020.5
-  - scipy=1.6
-  - sigtools
-  - sphinx
-  # Optional
-  # Development
-  - astroid
-  - bandit
-  - black=19.10b0
-  - codacy-coverage
-  - filelock
-  - flake8
-  - isort=5.*
-  - mock
-  - mypy
-  - pytest
-  - pytest-cov
-  - pytest-xdist
-  - safety
-  - sphinx-autodoc-typehints
diff --git a/envs/latest.yml b/envs/latest.yml
new file mode 100644
index 0000000000..ae82212e3f
--- /dev/null
+++ b/envs/latest.yml
@@ -0,0 +1,44 @@
+# This environment is intended to provide all optional dependencies and
+# use as-recent-as-possible versions with minimal pinning.
+# Acceptance test failures due to different output values are OK.
+# Unit test failures are not OK - pinning should be used where needed to
+# make the unit tests pass.
+name: improver_latest
+channels:
+  - conda-forge
+dependencies:
+  - python=3
+  # Required
+  - cartopy<0.20
+  - cftime<1.5
+  - cf-units=2.1.5
+  - clize
+  - dask
+  - iris>=3.0
+  - netCDF4
+  - numpy<1.21
+  - pytz
+  - scipy
+  - sigtools
+  - sphinx
+  # Optional
+  - python-stratify
+  - statsmodels
+  - numba
+  - pysteps
+  - timezonefinder
+  # Development
+  - astroid
+  - bandit
+  - black
+  - codacy-coverage
+  - filelock
+  - flake8
+  - isort
+  - mock
+  - mypy
+  - pytest
+  - pytest-cov
+  - pytest-xdist
+  - safety
+  - sphinx-autodoc-typehints
diff --git a/improver/metadata/amend.py b/improver/metadata/amend.py
index 9d6a916187..28a3a30a9e 100644
--- a/improver/metadata/amend.py
+++ b/improver/metadata/amend.py
@@ -29,10 +29,9 @@
 # ARISING IN ANY WAY OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE
 # POSSIBILITY OF SUCH DAMAGE.
 """Module containing utilities for modifying cube metadata"""
-from datetime import datetime
+from datetime import datetime, timedelta, timezone
 from typing import Any, Dict, List, Union
 
-from dateutil import tz
 from iris.cube import Cube, CubeList
 
 from improver.metadata.constants.mo_attributes import (
@@ -94,8 +93,9 @@ def set_history_attribute(cube: Cube, value: str, append: bool = False) -> None:
             If True, add to the existing history rather than replacing the
             existing attribute.  Default is False.
     """
-    tzinfo = tz.tzoffset("Z", 0)
-    timestamp = datetime.strftime(datetime.now(tzinfo), "%Y-%m-%dT%H:%M:%S%Z")
+    timestamp = datetime.strftime(
+        datetime.now(timezone(timedelta(0), name="Z")), "%Y-%m-%dT%H:%M:%S%Z"
+    )
     new_history = "{}: {}".format(timestamp, value)
     if append and "history" in cube.attributes.keys():
         cube.attributes["history"] += "; {}".format(new_history)
diff --git a/improver/psychrometric_calculations/psychrometric_calculations.py b/improver/psychrometric_calculations/psychrometric_calculations.py
index e7ada36ec6..99d221c101 100644
--- a/improver/psychrometric_calculations/psychrometric_calculations.py
+++ b/improver/psychrometric_calculations/psychrometric_calculations.py
@@ -38,7 +38,6 @@
 from cf_units import Unit
 from iris.cube import Cube, CubeList
 from numpy import ndarray
-from stratify import interpolate
 
 import improver.constants as consts
 from improver import BasePlugin
@@ -587,6 +586,8 @@ def find_falling_level(
         Returns:
             Phase change level data asl.
         """
+        from stratify import interpolate
+
         # Create cube of heights above sea level for each height in
         # the wet bulb integral cube.
         asl = wb_int_data.copy()
diff --git a/improver_tests/acceptance/test_checksums.py b/improver_tests/acceptance/test_checksums.py
index 8cd01038c9..d08e8603cd 100644
--- a/improver_tests/acceptance/test_checksums.py
+++ b/improver_tests/acceptance/test_checksums.py
@@ -131,7 +131,10 @@ def test_checksums_sorted():
     This test doesn't depend on having the acceptance test data available, so
     can run with the unit tests.
     """
-    csum_paths = [str(path) for path in acc.acceptance_checksums().keys()]
+    try:
+        csum_paths = [str(path) for path in acc.acceptance_checksums().keys()]
+    except FileNotFoundError:
+        pytest.skip("no checksum file, likely due to package being installed")
     with temporary_sort_locale("C") as strcoll:
         csum_paths_sorted = sorted(csum_paths, key=functools.cmp_to_key(strcoll))
     assert csum_paths == csum_paths_sorted
diff --git a/improver_tests/acceptance/test_interpolate_using_difference.py b/improver_tests/acceptance/test_interpolate_using_difference.py
index 2cea0b8ea3..fd51706ec2 100644
--- a/improver_tests/acceptance/test_interpolate_using_difference.py
+++ b/improver_tests/acceptance/test_interpolate_using_difference.py
@@ -85,6 +85,7 @@ def test_filling_with_nearest_use(tmp_path):
     """Test filling masked areas using difference interpolation, in this case
     with a hole in the corner of the data that requires use of nearest
     neighbour interpolation."""
+    pytest.importorskip("stratify")
     kgo_dir = acc.kgo_root() / f"{CLI}/basic"
     kgo_path = kgo_dir / "sleet_rain_nearest_filled_kgo.nc"
     output_path = tmp_path / "output.nc"
diff --git a/improver_tests/acceptance/test_phase_change_level.py b/improver_tests/acceptance/test_phase_change_level.py
index 13a986f08d..035c721282 100644
--- a/improver_tests/acceptance/test_phase_change_level.py
+++ b/improver_tests/acceptance/test_phase_change_level.py
@@ -55,6 +55,7 @@ def test_phase_change(tmp_path, phase_type, kgo_name, horiz_interp):
         snow/sleet level
         sleet/rain level leaving below orography points unfilled.
     """
+    pytest.importorskip("stratify")
     kgo_dir = acc.kgo_root() / f"{CLI}/basic"
     kgo_name = "{}_kgo.nc".format(kgo_name)
     kgo_path = kgo_dir / kgo_name
diff --git a/improver_tests/generate_ancillaries/test_GenerateTimezoneMask.py b/improver_tests/generate_ancillaries/test_GenerateTimezoneMask.py
index 2a02423fa4..14aabf638c 100644
--- a/improver_tests/generate_ancillaries/test_GenerateTimezoneMask.py
+++ b/improver_tests/generate_ancillaries/test_GenerateTimezoneMask.py
@@ -89,11 +89,24 @@ def global_grid_fixture_360() -> Cube:
     return cube
 
 
+@pytest.fixture(name="europe_grid")
+def europe_grid_fixture() -> Cube:
+    data = np.zeros((10, 10), dtype=np.float32)
+    cube = set_up_variable_cube(
+        data,
+        name="template",
+        grid_spacing=1,
+        domain_corner=(45, -5),
+        attributes=GLOBAL_ATTRIBUTES,
+    )
+    return cube
+
+
 @pytest.fixture(name="uk_grid")
 def uk_grid_fixture() -> Cube:
     """UK grid template"""
 
-    data = np.zeros((21, 22), dtype=np.float32)
+    data = np.zeros((11, 11), dtype=np.float32)
     cube = set_up_variable_cube(
         data,
         name="template",
@@ -160,7 +173,7 @@ def test__get_coordinate_pairs(request, grid_fixture):
     expected_data = {
         "global_grid": [[-90.0, -180.0], [-90.0, -80.0], [90.0, 180.0]],
         "global_grid_360": [[-90.0, -180.0], [-90.0, -80.0], [90.0, 180.0]],
-        "uk_grid": [[44.517, -17.117], [45.548, -4.913], [62.026, 14.410]],
+        "uk_grid": [[44.517, -17.117], [45.548, -4.913], [54.263, -5.401]],
     }
 
     grid = request.getfixturevalue(grid_fixture)
@@ -245,7 +258,7 @@ def test__create_template_cube(request, grid_fixture, include_dst):
 
     expected = {
         "global_grid": {"shape": (19, 37), "attributes": GLOBAL_ATTRIBUTES},
-        "uk_grid": {"shape": (21, 22), "attributes": UK_ATTRIBUTES},
+        "uk_grid": {"shape": (11, 11), "attributes": UK_ATTRIBUTES},
     }
 
     # Set expected includes_daylight_savings attribute
@@ -309,50 +322,98 @@ def test__group_timezones_empty_group(timezone_mask):
 
 
 # Expected data for process tests
-
-# ungrouped
-GLOBAL_GRID = {"shape": (27, 19, 37), "min": -12 * 3600, "max": 14 * 3600}
-# grouped
-GLOBAL_GRID_GR = {"shape": (2, 19, 37), "min": -6 * 3600, "max": 6 * 3600}
-UK_GRID_GR = {"shape": (2, 21, 22), "min": -6 * 3600, "max": 6 * 3600}
-
 EXPECTED_TIME = {None: 1510286400, "20200716T1500Z": 1594911600}
+GROUPED_MIN_MAX = {"min": -6 * 3600, "max": 6 * 3600}
 EXPECTED = {
     "ungrouped": {
-        "global": {
-            None: {**GLOBAL_GRID, "data": np.array([1, 1, 1, 1, 1, 0, 1, 1, 1, 1])},
-            "20200716T1500Z": {**GLOBAL_GRID, "data": np.ones([10])},
-            "indices": (12, 2),
-        },
         "uk": {
             None: {
-                "shape": (4, 21, 22),
-                "min": -2 * 3600,
+                "shape": (3, 11, 11),
+                "min": -1 * 3600,
+                "max": 1 * 3600,
+                "data": np.array(
+                    [
+                        [0, 0, 0, 0, 0, 0, 0, 1, 1, 1, 1],
+                        [1, 1, 1, 1, 1, 1, 1, 0, 0, 0, 0],
+                        [1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1],
+                    ]
+                ),
+            },
+            "20200716T1500Z": {
+                "shape": (4, 11, 11),
+                "min": -1 * 3600,
+                "max": 2 * 3600,
+                "data": np.array(
+                    [
+                        [0, 0, 0, 0, 0, 0, 0, 1, 1, 1, 1],
+                        [1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 0],
+                        [1, 1, 1, 1, 1, 1, 1, 0, 0, 0, 1],
+                        [1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1],
+                    ],
+                ),
+            },
+        },
+        "europe": {
+            None: {
+                "shape": (2, 10, 10),
+                "min": 0,
                 "max": 1 * 3600,
-                "data": np.array([1, 1, 0, 0, 0, 1]),
+                "data": np.array(
+                    [[0, 0, 0, 0, 0, 0, 0, 0, 0, 0], [1, 1, 1, 1, 1, 1, 1, 1, 1, 1]]
+                ),
             },
             "20200716T1500Z": {
-                "shape": (5, 21, 22),
-                "min": -2 * 3600,
+                "shape": (3, 10, 10),
+                "min": 0,
                 "max": 2 * 3600,
-                "data": np.array([1, 1, 1, 1, 0, 1]),
+                "data": np.array(
+                    [
+                        [1, 0, 1, 1, 1, 1, 0, 0, 0, 0],
+                        [0, 1, 0, 0, 0, 0, 1, 1, 1, 1],
+                        [1, 1, 1, 1, 1, 1, 1, 1, 1, 1],
+                    ]
+                ),
             },
-            "indices": (2, 10),
         },
     },
     "grouped": {
-        "global": {
-            None: {**GLOBAL_GRID_GR, "data": np.array([0, 0, 0, 0, 0, 0, 1, 1, 1, 1])},
+        "uk": {
+            None: {
+                "shape": (2, 11, 11),
+                **GROUPED_MIN_MAX,
+                "data": np.array(
+                    [
+                        [0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0],
+                        [1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1],
+                    ]
+                ),
+            },
             "20200716T1500Z": {
-                **GLOBAL_GRID_GR,
-                "data": np.array([0, 0, 0, 0, 0, 1, 1, 1, 1, 1]),
+                "shape": (2, 11, 11),
+                **GROUPED_MIN_MAX,
+                "data": np.array(
+                    [
+                        [0, 0, 0, 0, 0, 0, 0, 1, 1, 1, 0],
+                        [1, 1, 1, 1, 1, 1, 1, 0, 0, 0, 1],
+                    ],
+                ),
             },
-            "indices": (0, 2),
         },
-        "uk": {
-            None: {**UK_GRID_GR, "data": np.array([0, 0, 0, 0, 0, 1])},
-            "20200716T1500Z": {**UK_GRID_GR, "data": np.array([0, 0, 1, 1, 0, 1])},
-            "indices": (0, 9),
+        "europe": {
+            None: {
+                "shape": (2, 10, 10),
+                **GROUPED_MIN_MAX,
+                "data": np.array(
+                    [[0, 0, 0, 0, 0, 0, 0, 0, 0, 0], [1, 1, 1, 1, 1, 1, 1, 1, 1, 1]]
+                ),
+            },
+            "20200716T1500Z": {
+                "shape": (2, 10, 10),
+                **GROUPED_MIN_MAX,
+                "data": np.array(
+                    [[1, 0, 1, 1, 1, 1, 0, 0, 0, 0], [0, 1, 0, 0, 0, 0, 1, 1, 1, 1]]
+                ),
+            },
         },
     },
 }
@@ -360,7 +421,7 @@ def test__group_timezones_empty_group(timezone_mask):
 
 @pytest.mark.parametrize("grouping", ["ungrouped", "grouped"])
 @pytest.mark.parametrize("time", [None, "20200716T1500Z"])
-@pytest.mark.parametrize("grid_fixture", ["global_grid", "global_grid_360", "uk_grid"])
+@pytest.mark.parametrize("grid_fixture", ["uk_grid", "europe_grid"])
 def test_process(request, grid_fixture, time, grouping):
     """Test that the process method returns cubes that take the expected form
     for different grids and different dates.
@@ -376,7 +437,6 @@ def test_process(request, grid_fixture, time, grouping):
 
     expected = EXPECTED[grouping][domain][time]
     expected_time = EXPECTED_TIME[time]
-    index = EXPECTED[grouping][domain]["indices"]
 
     grid = request.getfixturevalue(grid_fixture)
 
@@ -393,5 +453,14 @@ def test_process(request, grid_fixture, time, grouping):
         assert (
             result.coord("UTC_offset").bounds.dtype == TIME_COORDS["UTC_offset"].dtype
         )
+    # slice the first spatial dimension to moderate size of expected arrays
+    assert_array_equal(result.data[:, 9, :], expected["data"])
 
-    assert_array_equal(result.data[index][::4], expected["data"])
+    # check each spatial location in the UTC_offset dimension
+    zone_count = np.count_nonzero(result.data, axis=0)
+    if grouping == "grouped":
+        # grouped outputs have a single UTC_offset with a non-zero entry
+        assert_array_equal(zone_count, 1)
+    else:
+        # ungrouped outputs have a single UTC_offset with a zero entry
+        assert_array_equal(zone_count, expected["shape"][0] - 1)
diff --git a/improver_tests/psychrometric_calculations/test_PhaseChangeLevel.py b/improver_tests/psychrometric_calculations/test_PhaseChangeLevel.py
index 237e4b89b7..e061baafb3 100644
--- a/improver_tests/psychrometric_calculations/test_PhaseChangeLevel.py
+++ b/improver_tests/psychrometric_calculations/test_PhaseChangeLevel.py
@@ -34,6 +34,7 @@
 
 import iris
 import numpy as np
+import pytest
 from cf_units import Unit
 from iris.cube import CubeList
 from iris.tests import IrisTest
@@ -109,6 +110,7 @@ class Test_find_falling_level(IrisTest):
 
     def setUp(self):
         """Set up arrays."""
+        pytest.importorskip("stratify")
         self.wb_int_data = np.array(
             [
                 [[80.0, 80.0], [70.0, 50.0]],
@@ -449,7 +451,7 @@ def setUp(self):
         """Set up orography and land-sea mask cubes. Also create temperature,
         pressure, and relative humidity cubes that contain multiple height
         levels."""
-
+        pytest.importorskip("stratify")
         self.setup_cubes_for_process()
 
     def setup_cubes_for_process(self, spatial_grid="equalarea"):
diff --git a/setup.cfg b/setup.cfg
index 73a8026068..74a7564144 100644
--- a/setup.cfg
+++ b/setup.cfg
@@ -18,45 +18,15 @@ packages = find:
 setup_requires =
     setuptools >= 38.3.0
     setuptools_scm
-install_requires =
-    cartopy
-    cftime == 1.0.1
-    cf_units
-    clize
-    dask
-    netCDF4
-    numpy
-    python-dateutil
-    pytz
-    scipy >= 1.3.0, < 1.4.0
-    scitools-iris >= 3.0
-    sigtools
-    sphinx
-    statsmodels
-    stratify
+# Note: no install_requires run-time requirements are included here.
+# Requirements are expected to be provided through another method such as conda.
+# See envs directory at top level of repository.
 scripts = bin/improver
 
 [options.packages.find]
-exclude = improver_tests
-
-[options.extras_require]
-dev =
-    astroid
-    bandit
-    black == 19.10b0
-    codacy-coverage
-    filelock
-    isort == 5.*
-    mock
-    mypy
-    pytest
-    pytest-cov
-    safety
-    sphinx-autodoc-typehints
-full =
-    numba
-    pysteps == 1.3.2
-    timezonefinder
+exclude =
+    improver_tests
+    improver_tests.*
 
 [flake8]
 max-line-length = 100