Forecast.Solar predictions dip at midday, every day and are offset vs actuals by a few hours

Every single day the forecast has a dip at midday:

Location is Melbourne Australia and our weather definitely isn’t so routine as to only have cloud cover precisely at midday every day! I’ve removed and re-added the integration several times, but get the same result.
I really hope I’m not being dumb with the configuration information I’m entering.

Panels face 40deg East of North, so I enter “40”.
Panel angle is 20deg above the horizontal, so I enter “20”.
Lat/long and timezone are all correct.
Yet every day the forecast expects panel output to drop at midday.

Debug info:

  "data": {
    "entry": {
      "title": "Home",
      "data": {
        "latitude": "**REDACTED**",
        "longitude": "**REDACTED**"
      },
      "options": {
        "declination": 20,
        "azimuth": 40,
        "modules power": 10400,
        "damping": 0.0,
        "inverter_size": 8200
      }
    },
    "data": {
      "energy_production_today": 12280,
      "energy_production_tomorrow": 21459,
      "energy_current_hour": 1400,
      "power_production_now": 1400,
      "watts": {
        "2022-04-27T06:50:00+10:00": 0,
        "2022-04-27T07:25:00+10:00": 0,
        "2022-04-27T08:00:00+10:00": 1,
        "2022-04-27T09:00:00+10:00": 112,
        "2022-04-27T10:00:00+10:00": 430,
        "2022-04-27T11:30:00+10:00": 1741,
        "2022-04-27T12:00:00+10:00": 1876,
        "2022-04-27T13:00:00+10:00": 2036,
        "2022-04-27T14:00:00+10:00": 1970,
        "2022-04-27T15:00:00+10:00": 1747,
        "2022-04-27T16:00:00+10:00": 1400,
        "2022-04-27T17:00:00+10:00": 962,
        "2022-04-27T17:22:00+10:00": 198,
        "2022-04-27T17:44:00+10:00": 0,
        "2022-04-28T06:51:00+10:00": 0,
        "2022-04-28T07:26:00+10:00": 0,
        "2022-04-28T08:00:00+10:00": 0,
        "2022-04-28T09:00:00+10:00": 142,
        "2022-04-28T10:00:00+10:00": 723,
        "2022-04-28T11:30:00+10:00": 2201,
        "2022-04-28T12:00:00+10:00": 2635,
        "2022-04-28T13:00:00+10:00": 3330,
        "2022-04-28T14:00:00+10:00": 3626,
        "2022-04-28T15:00:00+10:00": 3587,
        "2022-04-28T16:00:00+10:00": 3117,
        "2022-04-28T17:00:00+10:00": 2162,
        "2022-04-28T17:21:00+10:00": 437,
        "2022-04-28T17:42:00+10:00": 0
      },
      "wh_days": {
        "2022-04-27T00:00:00": 12280,
        "2022-04-28T00:00:00": 21459
      },
      "wh_hours": {
        "2022-04-27T06:50:00+10:00": 0,
        "2022-04-27T07:25:00+10:00": 0,
        "2022-04-27T08:00:00+10:00": 1,
        "2022-04-27T09:00:00+10:00": 112,
        "2022-04-27T10:00:00+10:00": 430,
        "2022-04-27T11:30:00+10:00": 2611,
        "2022-04-27T12:00:00+10:00": 938,
        "2022-04-27T13:00:00+10:00": 2036,
        "2022-04-27T14:00:00+10:00": 1970,
        "2022-04-27T15:00:00+10:00": 1747,
        "2022-04-27T16:00:00+10:00": 1400,
        "2022-04-27T17:00:00+10:00": 962,
        "2022-04-27T17:22:00+10:00": 73,
        "2022-04-27T17:44:00+10:00": 0,
        "2022-04-28T06:51:00+10:00": 0,
        "2022-04-28T07:26:00+10:00": 0,
        "2022-04-28T08:00:00+10:00": 0,
        "2022-04-28T09:00:00+10:00": 142,
        "2022-04-28T10:00:00+10:00": 723,
        "2022-04-28T11:30:00+10:00": 3302,
        "2022-04-28T12:00:00+10:00": 1317,
        "2022-04-28T13:00:00+10:00": 3330,
        "2022-04-28T14:00:00+10:00": 3626,
        "2022-04-28T15:00:00+10:00": 3587,
        "2022-04-28T16:00:00+10:00": 3117,
        "2022-04-28T17:00:00+10:00": 2162,
        "2022-04-28T17:21:00+10:00": 153,
        "2022-04-28T17:42:00+10:00": 0
      }
    },
    "account": {
      "type": "public",
      "rate_limit": 12,
      "timezone": "Australia/Melbourne"
    }

Some dialogue over on the Forecast.Solar discord suggests that this dip and the offset by a few hours is a bit of a southern hemisphere thing. Will update here if I find more.

Just FYI I had this issue too. As I am also using a custom Solcast chart I didn’t ever bother sorting out what was happening.

Screenshot 2022-04-28 at 10-12-17 Overview – Home Assistant

The dip sould be fixed since last release.

Hi
I have the same issue, dip midday, also seems start and finish of production a bis shifted
Screen Shot 2022-05-29 at 10.10.14 pm

running 2022.5.5

I haven’t seen any updates come though, and the middle of the chart dip is still occurring.
The home assistant github for forecast.solar suggests no updates in the last 14 days, and that update was for translations.

My current install details:

“home_assistant”: {
“installation_type”: “Home Assistant OS”,
“version”: “2022.5.5”,
“dev”: false,
“hassio”: true,
“virtualenv”: false,
“python_version”: “3.9.9”,
“docker”: true,
“arch”: “x86_64”,
“timezone”: “Australia/Sydney”,
“os_name”: “Linux”,
“os_version”: “5.15.41”,
“supervisor”: “2022.05.3”,
“host_os”: “Home Assistant OS 8.1”,
“docker_version”: “20.10.14”,

forecast_solar==2.2.0

another Australian here - with the same precipitous dip in predicted production for the 1100 - 1200 hourly period. Keen to get this working to allow smarter control of my water heating and EV charging. I have very limited coding sklls but would be happy to help with troubleshooting if I can.

I’m on Home Assistant 2022.8.5 and still seeing this dip. Is that because Home Assistant still is using old code?