Each observation is compared to its historical baseline using a standardised departure (Z-score). The result is expressed as a temperature departure in °C and a percentile rank.
FORMULAZ = (T − μ) / σ
TObserved mean air temperature (°C)
μHistorical baseline mean (°C)
σHistorical standard deviation (°C)
DEPARTURET − μ (positive = warmer than average)
PERCENTILEΦ(Z) — cumulative normal distribution
WATCH|Z| > 1.28 · 90th %ile (ETCCDI TX90p)
MODERATE|Z| > 1.645 · 95th %ile (ETCCDI TX95p)
SIGNIFICANT|Z| > 2.326 · 99th %ile (IPCC extreme)
EXCEPTIONAL|Z| > 3.09 · 99.9th %ile (1-in-1,000)
Three baseline periods are available. All share the 1991–2020 WMO 30-year normal derived from PRISM 4km gridded climate data.
DAYSingle calendar day of year (DOY 1–366)
MONTHAll days within the calendar month pooled
DJFDec – Feb (Winter)
MAMMar – May (Spring)
JJAJun – Aug (Summer)
SONSep – Nov (Autumn)
Pooled baselines combine per-day statistics using: μ̄ = mean(μᵢ), σ̄² = mean(σᵢ² + μᵢ²) − μ̄²
Live and baseline data drawn from the following public sources.
POINT OBSCIMIS Station 174 · Los Angeles · CA Dept. of Water Resources
VARIABLEDaily mean air temperature · live API · raw
LAGD-0 (same day); falls back to the most recent finalized prior-day observation
SPATIAL LAGD-2 provisional · baseline precomputed (1991–2020 μ/σ)
BASELINEPRISM 1991–2020 · WMO 30-year normal
Point observations (CIMIS) represent a single location within the Los Angeles coastal plain. Spatial anomalies are interpolated 4km grid cells and may not reflect microclimatic variation.
D-0 CIMIS data is provisional and may be absent until finalized; the system falls back to the most recent finalized prior-day observation automatically. PRISM spatial data is typically available at D-2.
Anomaly thresholds (|Z| ≥ 1.645, 2.326, 3.09) are aligned with ETCCDI percentile definitions. They assume approximately normally distributed values, which is a reasonable but imperfect assumption for daily temperature extremes.
Vegetation/moisture layers depend on cloud-free Landsat overpasses. Early-month composites may have fewer observations and higher uncertainty. The baseline requires a one-time GEE export before this layer becomes active.
The Vegetation (MSAVI) and Moisture (NDMI) layers are derived from Landsat Collection 2, Level-2 Surface Reflectance imagery composited over the current calendar month.
MSAVI FORMULA(2·NIR + 1 − √((2·NIR+1)² − 8·(NIR−RED))) / 2
NDMI FORMULA(NIR − SWIR1) / (NIR + SWIR1)
PLATFORMGoogle Earth Engine (computational intermediary)
SENSOR STACKLandsat 5/7/8/9 · C02 L2 SR · 2003–2024
RESOLUTION500m (resampled from 30m)
BASELINE22-year monthly median (2003–2024)
COMPOSITECurrent-month median; falls back to prior month if <5 clear observations
MSAVI anomaly direction: positive Z = denser/greener than historical; negative Z = sparser/stressed. NDMI: positive Z = wetter canopy; negative Z = moisture deficit (elevated fire risk).
This layer shows whether California's rivers, reservoirs, and coastal waters are murkier than usual for this time of year. When sediment — ash, charred debris, eroded soil — washes into waterways after a fire, the water becomes less transparent and shifts color. The satellite detects that shift. Darker teal = more unusual sediment loading relative to the 22-year baseline for the same calendar month.
WHAT'S SHOWNSurface water turbidity anomaly — how much murkier the water is than historical norm
POST-FIRE SIGNALAsh, char, and eroded soil entering waterways — primary debris-flow and water quality indicator after fire
DATA WINDOWCurrent 16-day Landsat composite (most recent cloud-free satellite passes)
BASELINE22-year monthly median for this calendar month (2003–2024)
RESOLUTION4 km per pixel — regional signal; pixel edges overlap surrounding land