Enumerated Type Definitions
These Literal lists describe the available options to API functions.
data_timeseries
DataTimeseriesField = Literal['all', 'anom', 'anom_d', 'anom_ds', 'anom_qnt', 'anom_s', 'clim', 'stdv', 'trend', 'vals']
data_timeseries field options.
Field to plot. If returning a csv, clim returns only information about the climatological cycle, while vals or anom will return a historical timeseries since 1950.
anom: Total anomaly: departure from climatological normal (vals - clim).anom_d: Detrended anomaly component: anomaly with trend removed.anom_ds: Detrended and destandardized anomaly: normalized anomaly signal.anom_qnt: Anomaly expressed as a quantile (0-1) relative to historical distribution.anom_s: Standardized anomaly: anomaly divided by climatological standard deviation.clim: Climatological mean: long-term average for each day/week/month of the year.stdv: Climatological standard deviation: historical variability for each period.trend: Linear trend component: long-term directional change in the variable.vals: Raw ERA5 reanalysis values in absolute units.all: Returns all available fields.
DataTimeseriesFormat = Literal['csv', 'nc']
data_timeseries format options.
File format of the returned data.
nc: NetCDF4 binary format. Optimized for geospatial analysis with xarray.csv: Comma-Separated Values text format. Suitable for Pandas tables, spreadsheets, databases, and LLMs.
DataTimeseriesFrequency = Literal['3-monthly', 'daily', 'hourly', 'monthly', 'weekly']
data_timeseries frequency options.
Temporal resolution of historical values.
hourly: Hourly values.daily: Daily averages.weekly: Calendar-locked weekly aggregated values (Monday-Sunday).monthly: Calendar-locked monthly aggregated values.3-monthly: 3-month (quarterly/seasonal) aggregated values.
DataTimeseriesUnits = Literal['SI', 'US']
data_timeseries units options.
Unit system for returned data.
SI: Metric: °C for temperature, mm/day for precipitation, m/s for wind speed, etc.US: Imperial: °F for temperature, in/day for precipitation, mph for wind speed, etc.
DataTimeseriesVariable = Literal['cc', 'cdd', 'dewpoint', 'dhi', 'dni', 'evap', 'hdd', 'hgt10', 'hgt250', 'hgt500', 'mslp', 'olr', 'precip', 'rh', 'si', 'sm', 'snow', 'sss', 'sst', 'st', 'temp', 'tmax', 'tmin', 'tsi', 'wdir', 'wdir100', 'wgst', 'wpow', 'wspd', 'wspd100']
data_timeseries variable options.
Forecast variable.
cc: Cloud cover fraction (0-1).cdd: Cooling degree days (°C | °F). Cumulative measure of warm conditions requiring cooling.dewpoint: 2-meter dewpoint temperature (°C | °F).evap: Evaporation (mm/day). Negative values indicate evaporation, positive indicate condensation.hdd: Heating degree days (°C | °F). Cumulative measure of cold conditions requiring heating.hgt10: Geopotential height at 10 hPa (m | ft).hgt250: Geopotential height at 250 hPa (m | ft).hgt500: Geopotential height at 500 hPa (m | ft). Atmospheric pressure level indicator.mslp: Mean sea level pressure (mbar).olr: Outgoing longwave radiation (W/m²).precip: Total precipitation (mm/day | in/day). Liquid and frozen water equivalent.rh: Relative humidity (0-1). Ratio of actual to saturation vapor pressure.si: Sea ice cover fraction (0-1).sm: Soil moisture in upper 10 cm (m³/m³). Volumetric water content.snow: Snow depth (m water equivalent).sss: Sea surface salinity (psu).sst: Sea surface temperature (°C | °F).st: Soil temperature in upper 10 cm (°C | °F).temp: 2-meter air temperature (°C | °F). Average temperature forecast.tmax: Daily maximum 2-meter temperature (°C | °F).tmin: Daily minimum 2-meter temperature (°C | °F).tsi: Total solar irradiance / insolation (W/m² hourly | kWh/m²/day daily).wdir: 10-meter wind direction (degrees). True direction from which wind is blowing.wdir100: 100-meter wind direction (degrees). True direction from which wind is blowing.wgst: Maximum 10-meter wind gust (m/s | mph).wpow: Wind power density at 10 meters (W/m²).wspd: 10-meter wind speed (m/s | mph).wspd100: 100-meter wind speed (m/s | mph). Useful for wind energy applications.dhi: Diffuse horizontal irradiance (W/m²). Solar radiation from the sky excluding direct sunlight.dni: Direct normal irradiance (W/m²). Direct solar radiation perpendicular to surface.
DataTimeseriesWeights = Literal['equal', 'population', 'solar_capacity', 'wind_capacity']
data_timeseries weights options.
Weighting scheme used to aggregate data across grid points within a shapefile region.
If unspecified, no weighted mean will be calculated and the function will return values for all available gridpoints in the polygon.
population: Weight each grid point by population. Useful when calculating heating/cooling demand.equal: Each grid point within the shapefile receives uniform weight.wind_capacity: Weight each grid point by total nameplate capacity of wind turbines. USA only.solar_capacity: Weight each grid point by total nameplate capacity of grid-scale solar resources. Does not include smaller residential solar. USA only.
downscale
DownscaleFormat = Literal['nc']
downscale format options.
File format of the returned data.
nc: NetCDF4 binary format. Optimized for geospatial analysis with xarray.
DownscaleFrequency = Literal['daily', 'hourly']
downscale frequency options.
Temporal resolution of downscaled forecast values.
daily: Daily downscaled values.hourly: Hourly downscaled values.
DownscaleModel = Literal['blend']
downscale model options.
Forecast model.
DownscaleReferenceClim = Literal['10_yr', '30_yr', '5_yr', 'salient']
downscale reference_clim options.
Reference climatology period used to calculate forecast anomalies, tercile boundaries, and skill score baselines.
salient: Salient Climatology: Salient's proprietary climatological reference optimized for anomaly calculations and tercile boundaries. Provides the most accurate baseline for evaluating Salient forecast skill.30_yr: 30-year Climatology (1991-2020): NOAA's standard 30-year historical average with temporal smoothing. Updated every 10 years following scientific convention. Use this for anomaly calculations that match climatological forecast models.10_yr: 10-year Normal: Rolling 10-year straight average, updated annually. Captures more recent climate trends and variability. Useful for benchmarking forecast skill against recent climate conditions.5_yr: 5-year Normal: Rolling 5-year straight average, updated annually. Most responsive to recent climate change and short-term trends. Best for applications requiring the most current climate baseline.
DownscaleUnits = Literal['SI', 'US']
downscale units options.
Unit system for returned data.
SI: Metric: °C for temperature, mm/day for precipitation, m/s for wind speed, etc.US: Imperial: °F for temperature, in/day for precipitation, mph for wind speed, etc.
DownscaleVariables = Literal['cc', 'cdd', 'dewpoint', 'dhi', 'dni', 'evap', 'gdd', 'hdd', 'heat_index', 'hgt10', 'hgt250', 'hgt500', 'mslp', 'olr', 'precip', 'rh', 'si', 'sm', 'snow', 'sp', 'sss', 'sst', 'st', 'temp', 'tmax', 'tmin', 'tsi', 'wdir', 'wdir100', 'wgst', 'wind_chill', 'wpow', 'wspd', 'wspd100']
downscale variables options.
Forecast variable.
cdd: Cooling degree days (°C | °F). Cumulative measure of warm conditions requiring cooling.hdd: Heating degree days (°C | °F). Cumulative measure of cold conditions requiring heating.mslp: Mean sea level pressure (mbar).precip: Total precipitation (mm/day | in/day). Liquid and frozen water equivalent.rh: Relative humidity (0-1). Ratio of actual to saturation vapor pressure.sm: Soil moisture in upper 10 cm (m³/m³). Volumetric water content.snow: Snow depth (m water equivalent).st: Soil temperature in upper 10 cm (°C | °F).temp: 2-meter air temperature (°C | °F). Average temperature forecast.tsi: Total solar irradiance / insolation (W/m² hourly | kWh/m²/day daily).wdir: 10-meter wind direction (degrees). True direction from which wind is blowing.wdir100: 100-meter wind direction (degrees). True direction from which wind is blowing.wgst: Maximum 10-meter wind gust (m/s | mph).wpow: Wind power density at 10 meters (W/m²).wspd: 10-meter wind speed (m/s | mph).wspd100: 100-meter wind speed (m/s | mph). Useful for wind energy applications.dewpoint: 2-meter dewpoint temperature (°C | °F).hgt10: Geopotential height at 10 hPa (m | ft).hgt500: Geopotential height at 500 hPa (m | ft). Atmospheric pressure level indicator.olr: Outgoing longwave radiation (W/m²).tmax: Daily maximum 2-meter temperature (°C | °F).tmin: Daily minimum 2-meter temperature (°C | °F).cc: Cloud cover fraction (0-1).evap: Evaporation (mm/day). Negative values indicate evaporation, positive indicate condensation.hgt250: Geopotential height at 250 hPa (m | ft).si: Sea ice cover fraction (0-1).sss: Sea surface salinity (psu).sst: Sea surface temperature (°C | °F).heat_index: Heat index at 2 meters (°C | °F). Apparent temperature combining heat and humidity.wind_chill: Wind chill at 2 meters (°C | °F). Apparent temperature combining cold and wind.gdd: Growing degree days. Measure of heat accumulation for agricultural applications.dhi: Diffuse horizontal irradiance (W/m²). Solar radiation from the sky excluding direct sunlight.dni: Direct normal irradiance (W/m²). Direct solar radiation perpendicular to surface.
DownscaleWeights = Literal['equal', 'population', 'solar_capacity', 'wind_capacity']
downscale weights options.
Weighting scheme used to aggregate data across grid points within a shapefile region.
If unspecified, no weighted mean will be calculated and the function will return values for all available gridpoints in the polygon.
population: Weight each grid point by population. Useful when calculating heating/cooling demand.equal: Each grid point within the shapefile receives uniform weight.wind_capacity: Weight each grid point by total nameplate capacity of wind turbines. USA only.solar_capacity: Weight each grid point by total nameplate capacity of grid-scale solar resources. Does not include smaller residential solar. USA only.
forecast_timeseries
ForecastTimeseriesField = Literal['anom', 'anom_ens', 'terciles', 'vals', 'vals_ens']
forecast_timeseries field options.
Type of data to return, typically a quantile-based probability distribution or a set of ensemble trajectories. N=31 to 300, depending on model.
vals: Probability quantiles of forecast values in absolute units.anom: Probability quantiles representing the departure from climatological normal.vals_ens: Ensemble trajectories in absolute units.anom_ens: Ensemble trajectories representing departure from climatological normal.terciles: Tercile categories indicating above-normal, near-normal, or below-normal conditions.
ForecastTimeseriesFormat = Literal['csv', 'nc']
forecast_timeseries format options.
File format of the returned data.
nc: NetCDF4 binary format. Optimized for geospatial analysis with xarray.csv: Comma-Separated Values text format. Suitable for Pandas tables, spreadsheets, databases, and LLMs.
ForecastTimeseriesModel = Literal['blend', 'clim', 'ecmwf', 'ecmwf_ens', 'ecmwf_ens15', 'ecmwf_ens_calib', 'ecmwf_seas5', 'ecmwf_seas5_calib', 'gem', 'gfs', 'noaa_gefs', 'noaa_gefs_calib', 'noaa_gfs', 'norm_10yr', 'norm_5yr', 'salient_clim']
forecast_timeseries model options.
Forecast model.
clim: Climatological forecasts: Probabilistic distributions characterizing uncertainty around calculated climatological means based on historical climate patterns.gfs: Generic reference to NOAA GFS models. Seenoaa_gfsfor details.ecmwf: Generic reference to ECMWF models. Seeecmwf_ensandecmwf_seas5for specific model details.blend: Salient Blend: Proprietary multi-model ensemble that objectively combines Salient's AI models with properly calibrated dynamical models (GEFS and ECMWF). Available as probability distributions for weekly through quarterly timescales. Usedownscalefor daily ensembles.gem: Salient GemAI: Generative AI model. Produces native daily ensemble trajectories for days 1-126 ahead, aggregating to coarser timescales and probability distributions.salient_clim: Salient's proprietary climatological model providing full probabilistic distributions with enhanced uncertainty characterization.norm_5yr: 5-year rolling climatological normal calculated from the most recent 5 years of observations.norm_10yr: 10-year rolling climatological normal calculated from the most recent 10 years of observations.noaa_gfs: NOAA GFS - Global Forecast System: single-member weather model providing hourly forecasts for hours 1-120 (~5 days).noaa_gefs: NOAA GEFS - Global Ensemble Forecast System: 31-member ensemble providing probabilistic forecasts for days 1-35.noaa_gefs_calib: Calibrated NOAA GEFS: GEFS ensemble with statistical post-processing to reduce systematic biases and improve reliability.ecmwf_ens: ECMWF ENS - Extended Range: European Centre ensemble model providing 101 ensemble members for 46 days of forecasts.ecmwf_ens_calib: Calibrated ECMWF ENS Extended Range: ECMWF ensemble with statistical calibration to improve forecast reliability and reduce biases.ecmwf_seas5: ECMWF SEAS5 - Seasonal forecast system: Ensemble covering months 1-3 and quarters 1-2ecmwf_seas5_calib: Calibrated ECMWF SEAS5: ECMWF seasonal model with post-processing for improved skill and reliability at seasonal timescales.ecmwf_ens15: ECMWF 15-day forecast
ForecastTimeseriesReferenceClim = Literal['10_yr', '30_yr', '5_yr', 'salient']
forecast_timeseries reference_clim options.
Reference climatology period used to calculate forecast anomalies, tercile boundaries, and skill score baselines.
salient: Salient Climatology: Salient's proprietary climatological reference optimized for anomaly calculations and tercile boundaries. Provides the most accurate baseline for evaluating Salient forecast skill.30_yr: 30-year Climatology (1991-2020): NOAA's standard 30-year historical average with temporal smoothing. Updated every 10 years following scientific convention. Use this for anomaly calculations that match climatological forecast models.10_yr: 10-year Normal: Rolling 10-year straight average, updated annually. Captures more recent climate trends and variability. Useful for benchmarking forecast skill against recent climate conditions.5_yr: 5-year Normal: Rolling 5-year straight average, updated annually. Most responsive to recent climate change and short-term trends. Best for applications requiring the most current climate baseline.
ForecastTimeseriesTimescale = Literal['all', 'daily', 'hourly', 'long-range', 'monthly', 'quarterly', 'seasonal', 'sub-seasonal', 'weekly']
forecast_timeseries timescale options.
Temporal resolution and lookahead range of the forecast.
all: Returns all available timescales for the selected model. Timescale availability varies by model.hourly: NOAA GFS hourly resolution covers hours 1-120 (~5 days) and initializes at 00z/06z/12z/18z.daily: Day-by-day ensemble forecasts, 35/46/126 days into the future (depending on model). Updates daily.sub-seasonal: Rolling weekly forecasts covering weeks 1-5 ahead of forecast_date. Updates daily.seasonal: Rolling monthly forecasts covering 3x30 day periods ahead of forecast_date. Updates daily or weekly depending on model.long-range: Rolling quarterly forecasts covering quarters 1-4 ahead for 12 months of predictability. Updates monthly on the 15th for theblendmodel.weekly: Calendar-aligned weekly forecasts covering weeks 1-4 or 5, starting each Monday and ending Sunday.monthly: Calendar-aligned monthly forecasts covering 2-3 calendar months ahead, starting on day 1 of the month.quarterly: Calendar-aligned seasonal forecasts covering seasons 1-3 or 4 (e.g., DJF, MAM, JJA, SON).
ForecastTimeseriesUnits = Literal['SI', 'US']
forecast_timeseries units options.
Unit system for returned data.
SI: Metric: °C for temperature, mm/day for precipitation, m/s for wind speed, etc.US: Imperial: °F for temperature, in/day for precipitation, mph for wind speed, etc.
ForecastTimeseriesVariable = Literal['cc', 'cdd', 'dewpoint', 'evap', 'hdd', 'heat_index', 'hgt10', 'hgt250', 'hgt500', 'mslp', 'olr', 'precip', 'rh', 'si', 'sm', 'snow', 'sp', 'sss', 'sst', 'st', 'temp', 'tmax', 'tmin', 'tsi', 'wdir', 'wdir100', 'wgst', 'wind_chill', 'wpow', 'wspd', 'wspd100']
forecast_timeseries variable options.
Forecast variable.
cdd: Cooling degree days (°C | °F). Cumulative measure of warm conditions requiring cooling.hdd: Heating degree days (°C | °F). Cumulative measure of cold conditions requiring heating.mslp: Mean sea level pressure (mbar).precip: Total precipitation (mm/day | in/day). Liquid and frozen water equivalent.rh: Relative humidity (0-1). Ratio of actual to saturation vapor pressure.sm: Soil moisture in upper 10 cm (m³/m³). Volumetric water content.snow: Snow depth (m water equivalent).st: Soil temperature in upper 10 cm (°C | °F).temp: 2-meter air temperature (°C | °F). Average temperature forecast.tsi: Total solar irradiance / insolation (W/m² hourly | kWh/m²/day daily).wdir: 10-meter wind direction (degrees). True direction from which wind is blowing.wdir100: 100-meter wind direction (degrees). True direction from which wind is blowing.wgst: Maximum 10-meter wind gust (m/s | mph).wpow: Wind power density at 10 meters (W/m²).wspd: 10-meter wind speed (m/s | mph).wspd100: 100-meter wind speed (m/s | mph). Useful for wind energy applications.dewpoint: 2-meter dewpoint temperature (°C | °F).hgt10: Geopotential height at 10 hPa (m | ft).hgt500: Geopotential height at 500 hPa (m | ft). Atmospheric pressure level indicator.olr: Outgoing longwave radiation (W/m²).tmax: Daily maximum 2-meter temperature (°C | °F).tmin: Daily minimum 2-meter temperature (°C | °F).cc: Cloud cover fraction (0-1).evap: Evaporation (mm/day). Negative values indicate evaporation, positive indicate condensation.hgt250: Geopotential height at 250 hPa (m | ft).si: Sea ice cover fraction (0-1).sss: Sea surface salinity (psu).sst: Sea surface temperature (°C | °F).heat_index: Heat index at 2 meters (°C | °F). Apparent temperature combining heat and humidity.wind_chill: Wind chill at 2 meters (°C | °F). Apparent temperature combining cold and wind.
ForecastTimeseriesWeights = Literal['equal', 'population', 'solar_capacity', 'wind_capacity']
forecast_timeseries weights options.
Weighting scheme used to aggregate data across grid points within a shapefile region.
If unspecified, no weighted mean will be calculated and the function will return values for all available gridpoints in the polygon.
population: Weight each grid point by population. Useful when calculating heating/cooling demand.equal: Each grid point within the shapefile receives uniform weight.wind_capacity: Weight each grid point by total nameplate capacity of wind turbines. USA only.solar_capacity: Weight each grid point by total nameplate capacity of grid-scale solar resources. Does not include smaller residential solar. USA only.
geo
GeoFormat = Literal['csv', 'nc']
geo format options.
File format of the returned data.
csv: Comma-Separated Values text format. Suitable for Pandas tables, spreadsheets, databases, and LLMs.nc: NetCDF4 binary format. Optimized for geospatial analysis with xarray.
GeoResolution = Literal[0.0625, 0.125, 0.25]
geo resolution options.
Spatial resolution of extracted geodata in the polygon, when used with shapefile.
0.25: 0.25° (~28 km). Native ERA5 resolution.0.125: 0.125° (~14 km). 2x oversampled.0.0625: 0.0625° (~7 km). 4x oversampled.
GeoVariables = Literal['bdod', 'cec', 'cfvo', 'clay', 'elevation', 'fal', 'lai_hv', 'lai_lv', 'lulc_bgc', 'lulc_bgc_per', 'lulc_igbp', 'lulc_igbp_per', 'lulc_lai', 'lulc_lai_per', 'lulc_land_water', 'lulc_land_water_per', 'lulc_lccs_cover', 'lulc_lccs_cover_per', 'lulc_lccs_hydro', 'lulc_lccs_hydro_per', 'lulc_lccs_use', 'lulc_lccs_use_per', 'lulc_pft', 'lulc_pft_per', 'lulc_umd', 'lulc_umd_per', 'nitrogen', 'ocd', 'ocs', 'phh2o', 'pop_density', 'population', 'sand', 'silt', 'slope', 'soc', 'solar_capacity', 'solar_capacity_0_axis', 'solar_capacity_1_axis', 'solar_capacity_2_axis', 'wind_capacity', 'wind_elev', 'wind_hub_height', 'wind_turbine_ct', 'wv0010', 'wv0033', 'wv1500']
geo variables options.
Comma-separated list of geospatial variables to extract from the GeoData Archive.
bdod: Bulk Density (Oven Dry) (g cm**-3).cec: Cation exchange capacity of the soil (meq 100g**-1).cfvo: Volumetric fraction of coarse fragments (> 2 mm) (cm3 100**cm-3 (% vol)).clay: Proportion of clay particles (< 0.002 mm) in the fine earth fraction (g 100g**-1 (% wt)).elevation: Elevation (m).fal: Forecast albedo ((0 - 1)).lai_hv: Leaf area index high vegetation (m2 m-2).lai_lv: Leaf area index low vegetation (m2 m-2).lulc_bgc: BGC Land Cover Class (unitless).lulc_bgc_per: Percent BGC Land Cover Class (percent).lulc_igbp: IGBP Land Cover Class (unitless).lulc_igbp_per: Percent IGBP Land Cover Class (percent).lulc_lai: LAI Land Cover Class (unitless).lulc_lai_per: Percent LAI Land Cover Class (percent).lulc_land_water: MOD44W Land/Water Classes (unitless).lulc_land_water_per: Percent MOD44W Land/Water Classes (percent).lulc_lccs_cover: FAO-LCCS Land Cover Class (unitless).lulc_lccs_cover_per: Percent FAO-LCCS Land Cover Class (percent).lulc_lccs_hydro: FAO-LCCS Surface Hydrology Class (unitless).lulc_lccs_hydro_per: Percent FAO-LCCS Surface Hydrology Class (percent).lulc_lccs_use: FAO-LCCS Land Use Class (unitless).lulc_lccs_use_per: Percent FAO-LCCS Land Use Class (percent).lulc_pft: PFT Land Cover Class (unitless).lulc_pft_per: Percent PFT Land Cover Class (percent).lulc_umd: UMD Land Cover Class (unitless).lulc_umd_per: Percent UMD Land Cover Class (percent).nitrogen: Total nitrogen (g kg**-1).ocd: Organic carbon density (kg m**3).ocs: Organic carbon stocks (kg m**2).phh2o: Soil pH measured in a soil-water solution (pH).pop_density: Population Density (persons km**-2).population: Population (persons).sand: Proportion of sand particles (> 0.05/0.063 mm) in the fine earth fraction (g 100g**-1 (% wt)).silt: Proportion of silt particles (≥ 0.002 mm and ≤ 0.05/0.063 mm) in the fine earth fraction (g 100g**-1 (% wt)).slope: Slope (percent).soc: Soil organic carbon content in the fine earth fraction (g kg**-1).solar_capacity: Total nameplate capacity of solar resources in each grid cell (MW).solar_capacity_0_axis: Total nameplate capacity of fixed-axis solar resources in each grid cell (MW).solar_capacity_1_axis: Total nameplate capacity of single-axis tracker solar resources in each grid cell (MW).solar_capacity_2_axis: Total nameplate capacity of two-axis tracker solar resources in each grid cell (MW).wind_capacity: Total nameplate capacity of wind turbines in each grid cell (MW).wind_elev: Capacity-weighted mean elevation of turbine locations (m).wind_hub_height: Capacity-weighted mean turbine hub height (m).wind_turbine_ct: Number of wind turbines in each grid cell (#).wv0010: Volumetric water content at 10kPa suction (cm3 100cm-3 (% vol)).wv0033: Volumetric water content at 33kPa suction (cm3 100cm-3 (% vol)).wv1500: Volumetric water content at 1500kPa suction (cm3 100cm-3 (% vol)).
hindcast_summary
HindcastSummaryFormat = Literal['csv']
hindcast_summary format options.
File format of the returned data.
csv: Comma-Separated Values text format. Suitable for Pandas tables, spreadsheets, databases, and LLMs.
HindcastSummaryInterpMethod = Literal['linear', 'nearest']
hindcast_summary interp_method options.
Interpolation method to use for the requested data not on native grid points.
nearest: Nearest neighbor interpolation. Returns the value from the closest grid point.linear: Bilinear interpolation. Weighted average of surrounding grid points.
HindcastSummaryMetric = Literal['crps', 'crps_skill_score', 'mae', 'mae_skill_score', 'rps', 'rps_skill_score']
hindcast_summary metric options.
Forecast skill score metric.
crps: Continuous Ranked Probability Score. Measures probabilistic forecast accuracy for continuous variables.rps: Ranked Probability Score. Measures probabilistic forecast accuracy for categorical outcomes.mae: Mean Absolute Error. Average magnitude of forecast errors.crps_skill_score: CRPS Skill Score. CRPS improvement relative toreferencemodel (1 = perfect, 0 = no improvement, negative = worse).rps_skill_score: RPS Skill Score. RPS improvement relative toreferencemodel (1 = perfect, 0 = no improvement, negative = worse).mae_skill_score: MAE Skill Score. MAE improvement relative toreferencemodel (1 = perfect, 0 = no improvement, negative = worse).
HindcastSummaryModel = Literal['blend', 'clim', 'ecmwf', 'ecmwf_ens', 'ecmwf_ens15', 'ecmwf_ens_calib', 'ecmwf_seas5', 'ecmwf_seas5_calib', 'gem', 'gfs', 'noaa_gefs', 'noaa_gefs_calib', 'noaa_gfs', 'norm_10yr', 'norm_5yr', 'salient_clim']
hindcast_summary model options.
Forecast model.
clim: Climatological forecasts: Probabilistic distributions characterizing uncertainty around calculated climatological means based on historical climate patterns.gfs: Generic reference to NOAA GFS models. Seenoaa_gfsfor details.ecmwf: Generic reference to ECMWF models. Seeecmwf_ensandecmwf_seas5for specific model details.blend: Salient Blend: Proprietary multi-model ensemble that objectively combines Salient's AI models with properly calibrated dynamical models (GEFS and ECMWF). Available as probability distributions for weekly through quarterly timescales. Usedownscalefor daily ensembles.gem: Salient GemAI: Generative AI model. Produces native daily ensemble trajectories for days 1-126 ahead, aggregating to coarser timescales and probability distributions.salient_clim: Salient's proprietary climatological model providing full probabilistic distributions with enhanced uncertainty characterization.norm_5yr: 5-year rolling climatological normal calculated from the most recent 5 years of observations.norm_10yr: 10-year rolling climatological normal calculated from the most recent 10 years of observations.noaa_gfs: NOAA GFS - Global Forecast System: single-member weather model providing hourly forecasts for hours 1-120 (~5 days).noaa_gefs: NOAA GEFS - Global Ensemble Forecast System: 31-member ensemble providing probabilistic forecasts for days 1-35.noaa_gefs_calib: Calibrated NOAA GEFS: GEFS ensemble with statistical post-processing to reduce systematic biases and improve reliability.ecmwf_ens: ECMWF ENS - Extended Range: European Centre ensemble model providing 101 ensemble members for 46 days of forecasts.ecmwf_ens_calib: Calibrated ECMWF ENS Extended Range: ECMWF ensemble with statistical calibration to improve forecast reliability and reduce biases.ecmwf_seas5: ECMWF SEAS5 - Seasonal forecast system: Ensemble covering months 1-3 and quarters 1-2ecmwf_seas5_calib: Calibrated ECMWF SEAS5: ECMWF seasonal model with post-processing for improved skill and reliability at seasonal timescales.ecmwf_ens15: ECMWF 15-day forecast
HindcastSummaryReference = Literal['blend', 'clim', 'ecmwf', 'ecmwf_ens', 'ecmwf_ens15', 'ecmwf_ens_calib', 'ecmwf_seas5', 'ecmwf_seas5_calib', 'gem', 'gfs', 'noaa_gefs', 'noaa_gefs_calib', 'noaa_gfs', 'norm_10yr', 'norm_5yr', 'salient_clim']
hindcast_summary reference options.
Forecast model.
clim: Climatological forecasts: Probabilistic distributions characterizing uncertainty around calculated climatological means based on historical climate patterns.gfs: Generic reference to NOAA GFS models. Seenoaa_gfsfor details.ecmwf: Generic reference to ECMWF models. Seeecmwf_ensandecmwf_seas5for specific model details.blend: Salient Blend: Proprietary multi-model ensemble that objectively combines Salient's AI models with properly calibrated dynamical models (GEFS and ECMWF). Available as probability distributions for weekly through quarterly timescales. Usedownscalefor daily ensembles.gem: Salient GemAI: Generative AI model. Produces native daily ensemble trajectories for days 1-126 ahead, aggregating to coarser timescales and probability distributions.salient_clim: Salient's proprietary climatological model providing full probabilistic distributions with enhanced uncertainty characterization.norm_5yr: 5-year rolling climatological normal calculated from the most recent 5 years of observations.norm_10yr: 10-year rolling climatological normal calculated from the most recent 10 years of observations.noaa_gfs: NOAA GFS - Global Forecast System: single-member weather model providing hourly forecasts for hours 1-120 (~5 days).noaa_gefs: NOAA GEFS - Global Ensemble Forecast System: 31-member ensemble providing probabilistic forecasts for days 1-35.noaa_gefs_calib: Calibrated NOAA GEFS: GEFS ensemble with statistical post-processing to reduce systematic biases and improve reliability.ecmwf_ens: ECMWF ENS - Extended Range: European Centre ensemble model providing 101 ensemble members for 46 days of forecasts.ecmwf_ens_calib: Calibrated ECMWF ENS Extended Range: ECMWF ensemble with statistical calibration to improve forecast reliability and reduce biases.ecmwf_seas5: ECMWF SEAS5 - Seasonal forecast system: Ensemble covering months 1-3 and quarters 1-2ecmwf_seas5_calib: Calibrated ECMWF SEAS5: ECMWF seasonal model with post-processing for improved skill and reliability at seasonal timescales.ecmwf_ens15: ECMWF 15-day forecast
HindcastSummaryRegion = Literal['africa', 'asia', 'brazil', 'europe', 'global', 'north-america', 'russia', 'south-america', 'south-pacific', 'usa']
hindcast_summary region options.
Accepts continents, countries, or U.S. states.
HindcastSummarySeason = Literal['DJF', 'JJA', 'MAM', 'SON', 'all']
hindcast_summary season options.
Season filter for hindcast validation metrics.
all: All seasons combined.DJF: December, January, February (Northern Hemisphere winter).MAM: March, April, May (Northern Hemisphere spring).JJA: June, July, August (Northern Hemisphere summer).SON: September, October, November (Northern Hemisphere autumn).
HindcastSummarySplitSet = Literal['all', 'test', 'validation']
hindcast_summary split_set options.
Selection of test or validation sets.
all: All data (combined test and validation sets).test: Test set only. Data held out during model training for final evaluation.validation: Validation set only. Data used for hyperparameter tuning during development.
HindcastSummaryTimescale = Literal['all', 'daily', 'long-range', 'seasonal', 'sub-seasonal']
hindcast_summary timescale options.
Temporal resolution and lookahead range of the forecast.
all: Returns all available timescales for the selected model. Timescale availability varies by model.daily: Day-by-day ensemble forecasts, 35/46/126 days into the future (depending on model). Updates daily.sub-seasonal: Rolling weekly forecasts covering weeks 1-5 ahead of forecast_date. Updates daily.seasonal: Rolling monthly forecasts covering 3x30 day periods ahead of forecast_date. Updates daily or weekly depending on model.long-range: Rolling quarterly forecasts covering quarters 1-4 ahead for 12 months of predictability. Updates monthly on the 15th for theblendmodel.
HindcastSummaryUnits = Literal['SI', 'US']
hindcast_summary units options.
Unit system for returned data.
SI: Metric: °C for temperature, mm/day for precipitation, m/s for wind speed, etc.US: Imperial: °F for temperature, in/day for precipitation, mph for wind speed, etc.
HindcastSummaryVariable = Literal['cc', 'cdd', 'dewpoint', 'hdd', 'heat_index', 'hgt500', 'mslp', 'precip', 'rh', 'temp', 'tmax', 'tmin', 'tsi', 'wgst', 'wind_chill', 'wspd', 'wspd100']
hindcast_summary variable options.
Forecast variable.
heat_index: Heat index at 2 meters (°C | °F). Apparent temperature combining heat and humidity.wgst: Maximum 10-meter wind gust (m/s | mph).wspd: 10-meter wind speed (m/s | mph).precip: Total precipitation (mm/day | in/day). Liquid and frozen water equivalent.tmin: Daily minimum 2-meter temperature (°C | °F).hdd: Heating degree days (°C | °F). Cumulative measure of cold conditions requiring heating.hgt500: Geopotential height at 500 hPa (m | ft). Atmospheric pressure level indicator.rh: Relative humidity (0-1). Ratio of actual to saturation vapor pressure.wind_chill: Wind chill at 2 meters (°C | °F). Apparent temperature combining cold and wind.temp: 2-meter air temperature (°C | °F). Average temperature forecast.wspd100: 100-meter wind speed (m/s | mph). Useful for wind energy applications.dewpoint: 2-meter dewpoint temperature (°C | °F).mslp: Mean sea level pressure (mbar).tsi: Total solar irradiance / insolation (W/m² hourly | kWh/m²/day daily).cc: Cloud cover fraction (0-1).tmax: Daily maximum 2-meter temperature (°C | °F).cdd: Cooling degree days (°C | °F). Cumulative measure of warm conditions requiring cooling.
met_observations
MetObservationsFormat = Literal['csv', 'nc']
met_observations format options.
File format of the returned data.
csv: Comma-Separated Values text format. Suitable for Pandas tables, spreadsheets, databases, and LLMs.nc: NetCDF4 binary format. Optimized for geospatial analysis with xarray.
MetObservationsVariables = Literal['precip', 'snow', 'temp', 'tmax', 'tmin', 'wdir', 'wspd']
met_observations variables options.
Comma-separated list of meteorological variables to retrieve from weather station observations.
precip: Daily total precipitation (mm).temp: Daily average temperature (°C).tmax: Daily maximum temperature (°C).tmin: Daily minimum temperature (°C).snow: Daily snowfall (mm).wdir: Daily average wind direction (degrees).wspd: Daily average wind speed (m/s).
met_stations
MetStationsFormat = Literal['csv']
met_stations format options.
File format of the returned data.
csv: Comma-Separated Values text format. Suitable for Pandas tables, spreadsheets, databases, and LLMs.