Setting up idealised mountain in FullGaussianGrid versus HEALPixGrid

Hi I have been running idealised mountain experiments. All good with FullGaussianGrid: the code below sets us a narrow ridge between 30N and 90N. However, when I switched to running with HEALPixGrid, the mountain is now only spanning latitudes 30N to 50N. Can someone give me some pointers as to how I should set this up for HEALPixGrid? Maybe, also, given how this is a very narrow ridge, I should not be approaching the problem this way, as it is bound to create spectral ringing. Here is the code: ```julia " For a narrow ridge h(λ, φ) = H · f(φ) · exp( −(a cosφ)² (λ−λ₀)² / 2σ² ) σ must be a Float64 and in km" H = 10000.0 # m σ = 1.0 # km λ₀ = 180 φ₀ = 60 f(φ) = boxcar_taper(φ, 30.0, 90.0, 1.0) d(λ, φ) = 6371.0 * abs(cos(deg2rad(φ)) * deg2rad(λ - λ₀)) # km set!(model, orography = (λ, φ) -> H * f(φ) * exp(-d(λ, φ)^2 / (2σ^2)), add=false) ```