Sanity Check and Filtering

First step of the data processing chain.

Field Boundaries

Field boundaries are determined using either the alpha algorithm (alphashape) with the alpha parameter equal to the number of coordinates:

import alphashape
alpha = len(position)
alpha_shape = alphashape.alphashape(points, alpha)

or, if the alpha shape algorithm is not successful, the ConvexHull algorithm:

from scipy.spatial import ConvexHull
hull = ConvexHull(points)
indices = hull.vertices

The dataset is discarded if the area inside the field boundaries is less than 200 square metres.

Dataset Size

The dataset is discarded if the total number of usable data points in the data set is less than 100.

Averaging

The median value in each spectrum (HH, VV, HV, and VH) is calculated.

Total power, polarized power, and unpolarized power are calculated:

temp_tot = (0.5 * (data_hh2 + data_vv2))
temp_q = 1.0 * (0.5 * (data_hh2 - data_vv2))
temp_pol = np.sqrt(data_u * data_u + temp_q * temp_q + data_v * data_v)
temp_unpol = temp_tot - temp_pol

Filtering

The data point is discarded if the total power for an integration exceeds 400 Kelvin or is less than 50 Kelvin.