GPP calibration

[AlexisRenchon]

Add GPP calibration (Loss: gpp alone, then gpp + lhf)

To consider:

• use air or canopy temperature in pmodel?
• suggested priors (do we want k_C3 and k_C4 beer lambert? would need to be added):

EKP.constrained_gaussian("pmodel_cstar", 0.41, 0.05, 0.2, 0.7),
EKP.constrained_gaussian("pmodel_β_c3", 146.0, 40.0, 50.0, 300.0),
EKP.constrained_gaussian("pmodel_β_c4", 16.222, 5.0, 5.0, 40.0),
EKP.constrained_gaussian("pmodel_α", 0.933, 0.02, 0.85, 0.999),
EKP.constrained_gaussian("moisture_stress_c", 0.27, 0.15, 0.05, 1.0),

• main has c3 + c4 weighted average, but currently linear interpolation (do we want nearest neighbor?)
• should we consider moisture_stress_c_C3 and moisture_stress_c_C4? (and what about cstar and alpha)

PR #1681 works but the code is outdated.