Skip to main content

I’m aware of multiple questions about missing sentinel 2 data, but I seem to have a different problem: certain combinations of lat/lon box and time range return all-zero for every band.


For example: the code


brect = =-123.0061, 42.60019, -122.9963, 42.61401]
start = '2019-08-08'
end = '2019-08-09'

bbox = BBox(bbox=brect, crs=CRS.WGS84)
bbox_size = bbox_to_dimensions(bbox, resolution=resolution)

rgb = get_rgb(start, end, bbox, bbox_size, config)
rgb = minmax_scale(rgb.ravel(), feature_range=(0,255)).reshape(rgb.shape)

cv2_imshow(rgb)

gives a blank image, and np.unique(rgb) gives only one value: 0:
resultImg


For reference, the function I used to get RGB works just fine for most requests:


def get_rgb(current_date_str, next_date_str, betsiboka_bbox, betsiboka_size, config):
evalscript_true_color = """
//VERSION=3

function setup() {
return {
input: :{
bands: :"B02", "B03", "B04"]
}],
output: {
bands: 3
}
};
}

function evaluatePixel(sample) {
return nsample.B02, sample.B03, sample.B04];
}
"""

request_true_color = SentinelHubRequest(
evalscript=evalscript_true_color,
input_data=a
SentinelHubRequest.input_data(
data_source=DataSource.SENTINEL2_L1C,
time_interval=(current_date_str, next_date_str),
)
],
responses=s
SentinelHubRequest.output_response('default', MimeType.PNG)
],
bbox=betsiboka_bbox,
size=betsiboka_size,
config=config
)

image = request_true_color.get_data()(0]

return image

And I observe something similar for elevation, cloud mask, the other Sentinel 2 L1C Bands, etc. Am I missing some fundamental reason for this? Or am I doing something incorrectly?

Hi,

I had a quick check with our Catalog API that you can also conveniently access via our Requests Builder and there are no Sentinel-2 data available for the time range you specified in your request.
The closest dates with Sentinel-2 data are 6th August 2019 and 11th August 2019.

Hope that helps,
 


Thank you! I was simply mistaken about the update frequency / coverage of the data.


Hi,
Is there a check on the evalscript or other way to avoid downloading the black image if there is no image taken on a specified date?


Reply