Add intersection process and update skin dashboard templates

[webb-ben]

Add intersection OGC API - process from pygeoapi-plugins in support of https://github.com/internetofwater/geoconnex.us/issues/184. The data passed to the pygeoapi/pygeoapi-skin-dashboard/templates/processes/process.html jinja template has changed so I had to update the skin-dashboard template (cc: @dblodgett-usgs potentially the same issue as https://labs.waterdata.usgs.gov/api/nldi/pygeoapi/processes/nldi-flowtrace).

Also made additional updates to html templates. The differences can be seen at the branch of this PR and master. A mostly-exhaustive list is of changes to the htmls representations are:

• Modification of _base.html and css to slightly improve the appearance on the navigation on mobile devices.
• Added more information to the landing page
• Include all links in collection specific landing pages
• Load full table in /items (as implemented in https://github.com/geopython/pygeoapi/pull/1148)
• Fix the OAProc HTML representation and implement OAProc Intersector

[webb-ben]

More Mobile:

[webb-ben]

Another note:

The intersector plugin json representation advertises a OAProc schema that includes enumeration of all available collections that an API client can use.

{
    ...
    "collection": {
        "title": "Feature Collection",
        "description": "Feature Collection",
        "keywords": [
            "OGC API",
            "collection"
        ],
        "schema": {
            "type": "string",
            "example": "hu02",
            "enum": [
                "hu02",
                "hu04",
                "hu06",
                "hu08",
                "hu10",
                "nat_aq",
                "principal_aq",
                "sec_hydrg_reg",
                "gages",
                "mainstems",
                "states",
                "counties",
                "aiannh",
                "cbsa",
                "ua10",
                "places",
                "pws",
                "dams"
            ]
        },
        "minOccurs": 1,
        "maxOccurs": 1,
        "metadata": null
    }
    ...
}

[ksonda]

Neat! So could take like the flow lines geojson response from nldi and get all the census places they run through. That would be a good vignette