"""Spectral indices calculations for Earth Engine images."""
from typing import Any, Dict, List, Union
import ee
# Common spectral indices definitions
[docs]
SPECTRAL_INDICES = {
# Vegetation indices
"NDVI": {"formula": "(NIR - RED) / (NIR + RED)", "category": "vegetation"},
"EVI": {
"formula": "2.5 * (NIR - RED) / (NIR + 6 * RED - 7.5 * BLUE + 1)",
"category": "vegetation",
},
"SAVI": {
"formula": "((NIR - RED) / (NIR + RED + L)) * (1 + L)",
"category": "vegetation",
},
"NDII": {"formula": "(NIR - SWIR1) / (NIR + SWIR1)", "category": "vegetation"},
"NDMI": {"formula": "(NIR - SWIR1) / (NIR + SWIR1)", "category": "vegetation"},
"NDSI": {"formula": "(GREEN - SWIR1) / (GREEN + SWIR1)", "category": "snow"},
"NDTI": {"formula": "(SWIR1 - SWIR2) / (SWIR1 + SWIR2)", "category": "vegetation"},
# Burn indices
"NBR": {"formula": "(NIR - SWIR2) / (NIR + SWIR2)", "category": "burn"},
"NBR2": {"formula": "(SWIR1 - SWIR2) / (SWIR1 + SWIR2)", "category": "burn"},
# Water indices
"NDWI": {"formula": "(GREEN - NIR) / (GREEN + NIR)", "category": "water"},
"MNDWI": {"formula": "(GREEN - SWIR1) / (GREEN + SWIR1)", "category": "water"},
# Urban/Built-up indices
"NDBI": {"formula": "(SWIR1 - NIR) / (SWIR1 + NIR)", "category": "urban"},
# Additional vegetation
"MSI": {"formula": "SWIR1 / NIR", "category": "vegetation"},
"GNDVI": {"formula": "(NIR - GREEN) / (NIR + GREEN)", "category": "vegetation"},
}
# Platform-specific band mappings
[docs]
BAND_MAPPING = {
"LANDSAT/LC08": {"BLUE": "B2", "GREEN": "B3", "RED": "B4", "NIR": "B5", "SWIR1": "B6", "SWIR2": "B7"},
"LANDSAT/LC09": {"BLUE": "B2", "GREEN": "B3", "RED": "B4", "NIR": "B5", "SWIR1": "B6", "SWIR2": "B7"},
"COPERNICUS/S2": {"BLUE": "B2", "GREEN": "B3", "RED": "B4", "NIR": "B8", "SWIR1": "B11", "SWIR2": "B12"},
"COPERNICUS/S2_SR": {
"BLUE": "B2",
"GREEN": "B3",
"RED": "B4",
"NIR": "B8",
"SWIR1": "B11",
"SWIR2": "B12",
},
}
# Index categories
[docs]
CATEGORIES = {
"vegetation": ["NDVI", "EVI", "SAVI", "NDII", "NDMI", "GNDVI", "MSI", "NDTI"],
"burn": ["NBR", "NBR2"],
"water": ["NDWI", "MNDWI"],
"snow": ["NDSI"],
"urban": ["NDBI"],
"all": list(SPECTRAL_INDICES.keys()),
}
def spectralIndices(
src: Union[ee.Image, ee.ImageCollection],
index: Union[str, List[str]] = "NDVI",
G: float = 2.5,
C1: float = 6.0,
C2: float = 7.5,
L: float = 1.0,
cexp: float = 1.16,
nexp: float = 2.0,
alpha: float = 0.1,
slope: float = 1.0,
intercept: float = 0.0,
gamma: float = 1.0,
omega: float = 2.0,
beta: float = 0.05,
k: float = 0.0,
fdelta: float = 0.581,
epsilon: float = 1.0,
kernel: str = "RBF",
sigma: Union[str, float] = "0.5 * (a + b)",
p: float = 2.0,
c: float = 1.0,
lambdaN: float = 858.5,
lambdaR: float = 645.0,
lambdaG: float = 555.0,
online: bool = False,
drop: bool = False,
**kwargs: Any,
) -> Union[ee.Image, ee.ImageCollection]:
"""Compute spectral indices for an image or image collection."""
indices_to_compute = _get_indices_to_compute(index)
is_image = isinstance(src, ee.image.Image)
ref = src if is_image else src.first()
dataset_id = ee.String(ref.get("system:id")).getInfo()
band_map = _get_band_mapping(dataset_id)
def compute_indices(img: ee.Image) -> ee.Image:
result = img
for idx_name in indices_to_compute:
if idx_name not in SPECTRAL_INDICES:
continue
try:
result = result.addBands(_compute_index_simple(img, idx_name, band_map, L))
except Exception:
continue
if drop:
index_names = [idx for idx in indices_to_compute if idx in SPECTRAL_INDICES]
result = result.select(img.bandNames().cat(ee.List(index_names)))
return result
if is_image:
return compute_indices(src)
return src.map(compute_indices)
def _get_indices_to_compute(index: Union[str, List[str]]) -> List[str]:
"""Parse index input and return list of indices to compute."""
if isinstance(index, list):
return index
return CATEGORIES.get(index, [index] if index in SPECTRAL_INDICES else ["NDVI"])
def _get_band_mapping(dataset_id: str) -> Dict[str, str]:
"""Get band names for the dataset based on dataset ID."""
for key, mapping in BAND_MAPPING.items():
if key in dataset_id:
return mapping
return {"BLUE": "B2", "GREEN": "B3", "RED": "B4", "NIR": "B8", "SWIR1": "B11", "SWIR2": "B12"}
def _compute_index_simple(img: ee.Image, index_name: str, bands: Dict[str, str], L: float = 1.0) -> ee.Image:
"""Compute index using band math."""
if index_name == "NDVI":
nir, red = img.select(bands["NIR"]), img.select(bands["RED"])
return nir.subtract(red).divide(nir.add(red)).rename("NDVI")
elif index_name == "EVI":
nir, red, blue = img.select(bands["NIR"]), img.select(bands["RED"]), img.select(bands["BLUE"])
return (
nir.subtract(red)
.divide(nir.add(red.multiply(6)).subtract(blue.multiply(7.5)).add(1))
.multiply(2.5)
.rename("EVI")
)
elif index_name == "SAVI":
nir, red = img.select(bands["NIR"]), img.select(bands["RED"])
return nir.subtract(red).divide(nir.add(red).add(L)).multiply(1 + L).rename("SAVI")
elif index_name in ("NDMI", "NDII"):
nir, swir1 = img.select(bands["NIR"]), img.select(bands["SWIR1"])
return nir.subtract(swir1).divide(nir.add(swir1)).rename(index_name)
elif index_name == "NBR":
nir, swir2 = img.select(bands["NIR"]), img.select(bands["SWIR2"])
return nir.subtract(swir2).divide(nir.add(swir2)).rename("NBR")
elif index_name == "NBR2":
swir1, swir2 = img.select(bands["SWIR1"]), img.select(bands["SWIR2"])
return swir1.subtract(swir2).divide(swir1.add(swir2)).rename("NBR2")
elif index_name == "NDWI":
green, nir = img.select(bands["GREEN"]), img.select(bands["NIR"])
return green.subtract(nir).divide(green.add(nir)).rename("NDWI")
elif index_name == "MNDWI":
green, swir1 = img.select(bands["GREEN"]), img.select(bands["SWIR1"])
return green.subtract(swir1).divide(green.add(swir1)).rename("MNDWI")
elif index_name == "NDBI":
swir1, nir = img.select(bands["SWIR1"]), img.select(bands["NIR"])
return swir1.subtract(nir).divide(swir1.add(nir)).rename("NDBI")
elif index_name == "NDSI":
green, swir1 = img.select(bands["GREEN"]), img.select(bands["SWIR1"])
return green.subtract(swir1).divide(green.add(swir1)).rename("NDSI")
elif index_name == "NDTI":
swir1, swir2 = img.select(bands["SWIR1"]), img.select(bands["SWIR2"])
return swir1.subtract(swir2).divide(swir1.add(swir2)).rename("NDTI")
elif index_name == "GNDVI":
nir, green = img.select(bands["NIR"]), img.select(bands["GREEN"])
return nir.subtract(green).divide(nir.add(green)).rename("GNDVI")
elif index_name == "MSI":
return img.select(bands["SWIR1"]).divide(img.select(bands["NIR"])).rename("MSI")
else:
raise ValueError(f"Index {index_name} not supported")
