Step 01 — Upload Field Data

Upload a CSV file containing your agrivoltaics field experiment data. Block and Site (AV/Control) columns are required.

Step 02 — Assign Variables

Required Variables
Optional Variables
Subset Data (Optional)
Block Structure
Are block names shared between AV and Control?

• Independent: AV and Control sites are spatially separated — assign different block names for each site (e.g., AV: A/B/C/D · Control: I/II/III/IV). Using a unique Plot ID per observation is also valid.

• Dependent: AV and Control plots are located within the same block — both share the same block names (e.g., both use I/II/III/IV).
Replicates
Does data have replicates? (more than one observation per Block or Block:Row)

Detected Model Formula

Field Layout Diagram

Developer Log

Posts tagged #agrivoltaics from today.agronomy4future.com

About

Why Agrivoltaics Needs Its Own Statistical Approach


Single Crop · Single Row ☀ Solar Panels Block 1 AV Block 1 = Plot 1 Control Block 1 = Plot 1 Block 2 AV Block 2 = Plot 2 Control Block 2 = Plot 2 Random effect: (1 | Site:Block) Fixed effect: Site ⚠ Treatment confounded with space → LMM accounts for this lmer(Yield ~ Site + (1|Site:Block))

Standard experimental designs — such as split-plot or randomized complete block design (RCBD) — are commonly applied in agrivoltaics research, but they are fundamentally misaligned with the actual structure of agrivoltaics experiments.


In a typical agrivoltaics study, the treatment (solar panel shading) is permanently fixed in space. Full randomization of treatment and control plots — as required by split-plot or RCBD assumptions — would demand two to four times the land area, and removing solar panels to create control conditions is simply not feasible. More critically, the treatment is perfectly confounded with space: any observed difference between AV and control plots may reflect spatial variation rather than the treatment effect itself.


Despite these constraints, many researchers continue to apply split-plot or RCBD models without accounting for the spatial structure inherent to agrivoltaics systems. This leads to misrepresentation of treatment effects and inflated or deflated estimates of variability.


Agrivoltaics Stat was developed to address this gap. By applying linear mixed models (LMM) that explicitly account for the spatial and structural constraints of agrivoltaics experiments, this tool ensures that statistical inference is both valid and appropriate for real field conditions.


Terminology AV Control Crop 1 Crop 3 Crop 2 Crop 1 Crop 2 Crop 1 Crop 1 Crop 3 Crop 3 Crop 2 Crop 3 Site Plot Crop 1 Crop 2 Crop 3 Block Row



R Packages:

lme4, lmerTest, emmeans, multcomp, multcompView, ggplot2


8 Layout Cases supported:

Case 1: Single crop · single row

Case 2: Single crop · multiple rows

Case 3: Single crop · single row · multiple seasons

Case 4: Single crop · multiple rows · multiple seasons

Case 5: Multiple cultivars · single row

Case 6: Multiple cultivars · multiple rows

Case 7: Multiple cultivars · single row · multiple seasons

Case 8: Multiple cultivars · multiple rows · multiple seasons