[REQ] Please add rain attribute to Smart Irrigation sensor

[Apacs]

Is your feature request related to a problem? Please describe.

I'm currently visualizing and analyzing the daily irrigation logic in Home Assistant using the Smart Irrigation integration. I'm trying to understand and verify how much precipitation was factored into the calculation on a given day. However, the precipitation value is not exposed as an attribute of the sensor.smart_irrigation_ entity. This makes it impossible to reference the precipitation value used internally during the ET/drainage/bucket computation — either for display in dashboards or to verify/debug calculations later.

Describe the solution you'd like

Please add precipitation (or rain) as an attribute to the Smart Irrigation sensor, just like et_value, bucket, drainage, etc. This would allow users to inspect and log the exact precipitation amount Smart Irrigation used during its last calculation cycle.

Example (in extra_state_attributes(self)):

"rain": self._rain,

Describe alternatives you've considered

I tried reconstructing the value manually using the OpenWeatherMap forecast/history or logs, but that leads to inaccuracies due to the dynamic hourly data and its limited availability.

Logging the debug output every night and manually parsing it works, but it's not ideal for visualization or long-term insights.

Using an additional precipitation sensor separately (e.g., from OWM History) shows actual rainfall but not necessarily what Smart Irrigation used internally at the calculation time.

Additional context

I'm building a dashboard showing daily ET₀, drainage, irrigation, and precipitation as recorded by Smart Irrigation, using statistics sensors. Precipitation is a critical part of the water balance model. Having it exposed would make Smart Irrigation more transparent, debuggable, and easier to validate over time.

Screenshot of current attributes (note: precipitation is missing):

[Apacs]

I have an additional observation and follow-up question:

Although I'm primarily requesting that the precipitation value (rain) be exposed as a zone attribute, I suspect that Smart Irrigation might not currently be considering this value in its internal delta calculation – even when precipitation is properly configured and present (e.g., via OpenWeatherMap or custom sensors).

In my setup, I ensured:

• A valid precipitation source is available (use_precipitation assumed true by default),
• The bucket was nearly full (30 mm) at calculation time,
• Nevertheless, a very high delta (29 mm) was recorded,
• The debug log showed no trace of precipitation handling (see attached log for July 27/28).

Based on this, I wonder:

• Should precipitation be subtracted from the ET value during delta computation by default?
• If yes: is it possible that precipitation is currently ignored unless continuous mode is enabled and a very specific sensor setup is used?
• If not: could you clarify how precipitation is supposed to influence the delta or the bucket, and under which conditions?

I may be misunderstanding something here – so clarification from your side would be greatly appreciated.
I'm are also happy to test any improvements or help isolate the problem further.

Attached is the full debug log: [20250727_SI.txt]

20250727_SI.txt

(The calculation ran at 23:00 and shows a sharp bucket refill and large delta without visible rain adjustment.)

[Apacs]

Additional findings based on several days of observation and logs

I would like to share further conclusions based on the observations and debug logs from the past few days, following up on my initial request regarding the missing rain attribute in the zone sensor.

Observation and context

I have been monitoring the Smart Irrigation behavior daily and visualizing values such as:

• ET₀ (evapotranspiration)
• drainage
• calculated delta
• zone bucket
• rain (as derived externally from OWMH)

The bucket has remained full (30 mm) over multiple days, even though there has been substantial rainfall. Despite this, the delta value has increased significantly during the nightly recalculations. This behavior raises questions about the internal logic.

What I expected to see

Based on my understanding of the delta calculation:

delta = ET₀ – drainage – rain

I assumed the following:

• When the bucket is full (30 mm), the delta should remain close to zero, unless significant evaporation occurs.
• If ET₀ is ~3.7 mm and no drainage occurred, delta should be approx. 3.7 minus rain.
• In one case, the delta jumped to over 54 mm despite ET₀ being only ~3.7 mm, and the bucket already being at 30 mm before the update.

This is mathematically inconsistent with the formula above — unless rain and/or previous state values are being ignored.

Clarifying the rain vs. delta situation

I understand now that the current zone sensor does not expose the rain value as an attribute. I also confirmed that ET₀ and delta are always equal in my case. This is visible in my statistics chart, where both lines overlap completely.

This would only make sense if both rain and drainage were considered zero internally. Since I explicitly disabled drainage and see no rain attribute, it appears that rain is not being considered in the delta calculation.

Example calculation

Let’s take a concrete example:

bucket before = 30 mm
drainage = 0
rain = 0 (as per SI, although external rain data exists)
ET₀ = 3.7 mm

delta = 3.7 – 0 – 0 = 3.7 mm

This would be fine. But in my logs, delta was reported as ~54 mm, which is not explainable via ET₀ alone. This suggests:

• Either the ET₀ was incorrectly scaled
• Or previous state values (like bucket or last_delta) were lost or ignored
• Or rain should have been subtracted, but wasn't

My expectations

• I expect the delta to reflect the true net loss from the soil bucket, including rainfall and drainage.
• If Smart Irrigation intentionally ignores rain for some reason, this should be documented clearly.
• If rain is used internally, then we expect to see it subtracted from delta — and ideally, also exposed via the rain attribute for transparency and diagnostics.

Is this still a feature request?

Given the observed behavior and its mismatch with expectations, I am no longer sure if this is a simple feature request or an actual bug in the precipitation handling.

Please let me know if:

• I misunderstand something about how delta is supposed to be calculated
• You expect the rain value to come from a different place (e.g. forecast only)
• The current behavior is intentional and has known limitations

Attachments (if helpful)

Let me know if you'd like me to attach:

• the raw debug logs from Smart Irrigation during calculation
• screenshots of the overlapping ET₀ and delta graphs
• OWMH-derived rain sensor values from the same timeframes

Thanks again for this great integration — I'm happy to help improve it through testing and feedback.