Is Solar Irradiance Really Rising in Brazil's Northeast? Cloud-Driven vs. Aerosol-Driven Brightening
ALLSKY_SFC_SW_DWN) in Petrolina-PE (Northeast Brazil solar corridor) rose +0.2558 MJ/m²/day per decade (+1.075 MJ/m²/day cumulative, 95% CI: [+0.1624, +0.3496], τ = +0.178, p = 9.384e-09) over 1984–2025. However, decomposing All-Sky from Clear-Sky irradiance reveals this gain is cloud-driven, not aerosol-driven: CLOUD_ATTEN_PCT fell -0.9647 %/decade (p = 2.833e-09) while CLRSKY_SFC_SW_DWN stayed statistically flat (p = 0.2901).
1. How can a planner distinguish durable solar resource gain from a cloud-cover artifact?
Direct Answer: By computing Cloud Attenuation Deficit (CLOUD_ATTEN_PCT = 1 - ALLSKY_SFC_SW_DWN / CLRSKY_SFC_SW_DWN) alongside raw clear-sky irradiance (CLRSKY_SFC_SW_DWN) using the same Seasonal Mann-Kendall engine.
If total sunlight reaching PV panels increases solely because cloud attenuation decreases while clear-sky sunlight remains stationary, grid planners face an important meteorological caveat: inter-decadal cloud regimes can oscillate with Atlantic/ENSO cycles and co-occur with regional drying. In Petrolina-PE, cloud reduction coincided with T2M warming of +0.3012 °C/dec (p = 4.663e-15), relative humidity decline of -0.8932 %/dec (p = 5.157e-05), and rainfall loss of -0.0721 mm/day/dec (p = 0.0291) — raising panel cell-temperature derating and solar Duck Curve midday curtailment pressure.
| Station | All-Sky Irradiance (ALLSKY) |
Cloud Attenuation (CLOUD_ATTEN_PCT) |
Clear-Sky Irradiance (CLRSKY) |
Physical Driver |
|---|---|---|---|---|
| Petrolina-PE (Caatinga NE) | +0.2558 MJ/m²/day/dec (p = 9.384e-09) | -0.9647 %/dec (p = 2.833e-09) | +0.0127 MJ/m²/day/dec (p = 0.2901, null) | Cloud thinning |
| Manaus-AM (Amazon) | +0.2413 MJ/m²/day/dec (p = 1.738e-10) | -0.8552 %/dec (p = 3.015e-08) | +0.0289 MJ/m²/day/dec (p = 0.7686, null) | Cloud thinning |
| Campinas-SP (Southeast) | +0.1621 MJ/m²/day/dec (p = 0.0008616) | +0.0378 %/dec (p = 0.9476, null) | +0.1897 MJ/m²/day/dec (p < 1e-16) | Aerosol / clear-sky shift |
2. Why does Campinas-SP behave oppositely to Petrolina-PE and Manaus-AM?
Direct Answer: In Campinas-SP, CLOUD_ATTEN_PCT shows no trend over 42 years (+0.0378 %/dec, p = 0.9476), whereas CLRSKY_SFC_SW_DWN rises sharply (+0.1897 MJ/m²/day/dec, τ = +0.322, p < 1e-16). This indicates atmospheric aerosol / column water-vapor transparency changes rather than cloud loss — pairing with severe compound drying (RH2M -1.6601 %/dec, PRECTOTCORR -0.2724 mm/day/dec) and wildfire smoke susceptibility tracked via NASA FIRMS (VIIRS_NOAA21_NRT, CMR C3264430167-LANCEMODIS).
3. Resposta em Português: A irradiação solar no Nordeste (Petrolina-PE) está realmente subindo?
Sim. Entre 1984 e 2025 (504 meses da NASA POWER), a irradiação solar total em Petrolina-PE aumentou +0,2558 MJ/m²/dia por década (+1,075 MJ/m²/dia acumulados, p = 9,384e-09). A atribuição do Cockpit-OMNI comprova que esse aumento decorre da queda de atenuação por nuvens (-0,9647 %/década, p = 2,833e-09), pois o céu limpo permaneceu sem tendência estatística (p = 0,2901). Simultaneamente, Petrolina aqueceu +0,3012 °C/década e perdeu umidade (-0,8932 %/década), reforçando a necessidade de armazenamento em baterias e gestão térmica da Curva do Pato solar.