feat: explicitly implement the `dtype` on `numpy` quantities (#1922) #1120

Job	Run time
test-windows (3.9, numpy>=1.21,<2.0.0)	57s
test-macos (3.9)	1m 45s
test-windows (3.10, numpy>=1.21,<2.0.0)	1m 30s
test-linux (3.9)	1m 34s
test-macos (3.9, numpy>=1.21,<2.0.0)	2m 5s
test-windows (3.11, numpy>=1.21,<2.0.0)	56s
test-linux (3.9, uncertainties==3.1.6)	1m 58s
test-macos (3.10)	2m 25s
test-linux (3.9, uncertainties>=3.1.6,<4.0.0)	1m 33s
test-macos (3.10, numpy>=1.21,<2.0.0)	2m 1s
test-linux (3.9, numpy>=1.21,<2.0.0)	1m 45s
test-macos (3.11)	1m 53s
test-linux (3.9, numpy>=1.21,<2.0.0, uncertainties==3.1.6)	2m 13s
test-macos (3.11, numpy>=1.21,<2.0.0)	2m 17s
test-linux (3.9, numpy>=1.21,<2.0.0, uncertainties>=3.1.6,<4.0.0)	2m 13s
test-linux (3.10)	1m 44s
test-linux (3.10, uncertainties==3.1.6)	1m 39s
test-linux (3.10, uncertainties>=3.1.6,<4.0.0)	1m 56s
test-linux (3.10, numpy>=1.21,<2.0.0)	1m 47s
test-linux (3.10, numpy>=1.21,<2.0.0, uncertainties==3.1.6)	2m 29s
test-linux (3.10, numpy>=1.21,<2.0.0, uncertainties>=3.1.6,<4.0.0)	2m 20s
test-linux (3.11)	1m 35s
test-linux (3.11, uncertainties==3.1.6)	1m 32s
test-linux (3.11, uncertainties>=3.1.6,<4.0.0)	1m 42s
test-linux (3.11, numpy>=1.21,<2.0.0)	1m 40s
test-linux (3.11, numpy>=1.21,<2.0.0, uncertainties==3.1.6)	2m 14s
test-linux (3.11, numpy>=1.21,<2.0.0, uncertainties>=3.1.6,<4.0.0)	2m 15s
test-linux (3.9, numpy, matplotlib==2.2.5)	3m 19s
test-linux (3.9, numpy, uncertainties, sparse xarray netCDF4 dask[complete]==2023.4.0 graphviz ba...	3m 48s
	57m 5s

Provide feedback