|
| 1 | +{ |
| 2 | + "cells": [ |
| 3 | + { |
| 4 | + "cell_type": "markdown", |
| 5 | + "id": "0", |
| 6 | + "metadata": {}, |
| 7 | + "source": [ |
| 8 | + "# Using the `gsw` toolbox to compute density" |
| 9 | + ] |
| 10 | + }, |
| 11 | + { |
| 12 | + "cell_type": "markdown", |
| 13 | + "id": "1", |
| 14 | + "metadata": {}, |
| 15 | + "source": [ |
| 16 | + "This tutorial shows how to use the [`gsw` toolbox](https://teos-10.github.io/GSW-Python/) (the Gibbs SeaWater Oceanographic Toolbox of TEOS-10) within Parcels to compute density from temperature and salinity fields. The `gsw` toolbox can be installed via `conda install gsw`." |
| 17 | + ] |
| 18 | + }, |
| 19 | + { |
| 20 | + "cell_type": "markdown", |
| 21 | + "id": "2", |
| 22 | + "metadata": {}, |
| 23 | + "source": [ |
| 24 | + "First, load the necessary libraries and the data:" |
| 25 | + ] |
| 26 | + }, |
| 27 | + { |
| 28 | + "cell_type": "code", |
| 29 | + "execution_count": null, |
| 30 | + "id": "3", |
| 31 | + "metadata": {}, |
| 32 | + "outputs": [], |
| 33 | + "source": [ |
| 34 | + "import numpy as np\n", |
| 35 | + "import xarray as xr\n", |
| 36 | + "\n", |
| 37 | + "import parcels\n", |
| 38 | + "\n", |
| 39 | + "# Load the CopernicusMarine data in the Agulhas region from the example_datasets\n", |
| 40 | + "example_dataset_folder = parcels.download_example_dataset(\n", |
| 41 | + " \"CopernicusMarine_data_for_Argo_tutorial\"\n", |
| 42 | + ")\n", |
| 43 | + "\n", |
| 44 | + "ds = xr.open_mfdataset(f\"{example_dataset_folder}/*.nc\", combine=\"by_coords\")\n", |
| 45 | + "\n", |
| 46 | + "# TODO check how we can get good performance without loading full dataset in memory\n", |
| 47 | + "ds.load() # load the dataset into memory\n", |
| 48 | + "\n", |
| 49 | + "fieldset = parcels.FieldSet.from_copernicusmarine(ds)" |
| 50 | + ] |
| 51 | + }, |
| 52 | + { |
| 53 | + "cell_type": "markdown", |
| 54 | + "id": "4", |
| 55 | + "metadata": {}, |
| 56 | + "source": [ |
| 57 | + "Now, define a custom Particle class that includes temperature, salinity, and density as variables, and create a ParticleSet with one particle at a known location:" |
| 58 | + ] |
| 59 | + }, |
| 60 | + { |
| 61 | + "cell_type": "code", |
| 62 | + "execution_count": null, |
| 63 | + "id": "5", |
| 64 | + "metadata": {}, |
| 65 | + "outputs": [], |
| 66 | + "source": [ |
| 67 | + "GSWParticle = parcels.Particle.add_variable(\n", |
| 68 | + " [\n", |
| 69 | + " parcels.Variable(\"temp\", dtype=np.float32, initial=np.nan),\n", |
| 70 | + " parcels.Variable(\"salt\", dtype=np.float32, initial=np.nan),\n", |
| 71 | + " parcels.Variable(\"density\", dtype=np.float32, initial=np.nan),\n", |
| 72 | + " ]\n", |
| 73 | + ")\n", |
| 74 | + "\n", |
| 75 | + "# Initiate one Argo float in the Agulhas Current\n", |
| 76 | + "pset = parcels.ParticleSet(\n", |
| 77 | + " fieldset=fieldset,\n", |
| 78 | + " pclass=GSWParticle,\n", |
| 79 | + " lon=[32],\n", |
| 80 | + " lat=[-31],\n", |
| 81 | + " depth=[200],\n", |
| 82 | + ")" |
| 83 | + ] |
| 84 | + }, |
| 85 | + { |
| 86 | + "cell_type": "markdown", |
| 87 | + "id": "6", |
| 88 | + "metadata": {}, |
| 89 | + "source": [ |
| 90 | + "Now (as the core part of this tutorial) define a custom kernel that uses the `gsw` toolbox to compute density from temperature and salinity:" |
| 91 | + ] |
| 92 | + }, |
| 93 | + { |
| 94 | + "cell_type": "code", |
| 95 | + "execution_count": null, |
| 96 | + "id": "7", |
| 97 | + "metadata": {}, |
| 98 | + "outputs": [], |
| 99 | + "source": [ |
| 100 | + "def ParcelsGSW(particles, fieldset):\n", |
| 101 | + " import gsw\n", |
| 102 | + "\n", |
| 103 | + " particles.temp = fieldset.thetao[particles]\n", |
| 104 | + " particles.salt = fieldset.so[particles]\n", |
| 105 | + " pressure = gsw.p_from_z(-particles.depth, particles.lat)\n", |
| 106 | + " particles.density = gsw.density.rho(particles.salt, particles.temp, pressure)" |
| 107 | + ] |
| 108 | + }, |
| 109 | + { |
| 110 | + "cell_type": "markdown", |
| 111 | + "id": "8", |
| 112 | + "metadata": {}, |
| 113 | + "source": [ |
| 114 | + "Finally, run the `ParcelsGSW` Kernel for one timestep and check (for Continuous Integration purposes) that the computed density is as expected:" |
| 115 | + ] |
| 116 | + }, |
| 117 | + { |
| 118 | + "cell_type": "code", |
| 119 | + "execution_count": null, |
| 120 | + "id": "9", |
| 121 | + "metadata": {}, |
| 122 | + "outputs": [], |
| 123 | + "source": [ |
| 124 | + "pset.execute(ParcelsGSW, runtime=np.timedelta64(1, \"s\"), dt=np.timedelta64(1, \"s\"))\n", |
| 125 | + "\n", |
| 126 | + "np.testing.assert_allclose(pset.density, [1026.8281], rtol=1e-5)\n", |
| 127 | + "\n", |
| 128 | + "print(\n", |
| 129 | + " f\"Temperature: {pset.temp[0]:.2f}, Salinity: {pset.salt[0]:.2f}, Density: {pset.density[0]:.2f}\"\n", |
| 130 | + ")" |
| 131 | + ] |
| 132 | + } |
| 133 | + ], |
| 134 | + "metadata": { |
| 135 | + "kernelspec": { |
| 136 | + "display_name": "parcels", |
| 137 | + "language": "python", |
| 138 | + "name": "python3" |
| 139 | + }, |
| 140 | + "language_info": { |
| 141 | + "codemirror_mode": { |
| 142 | + "name": "ipython", |
| 143 | + "version": 3 |
| 144 | + }, |
| 145 | + "file_extension": ".py", |
| 146 | + "mimetype": "text/x-python", |
| 147 | + "name": "python", |
| 148 | + "nbconvert_exporter": "python", |
| 149 | + "pygments_lexer": "ipython3", |
| 150 | + "version": "3.13.5" |
| 151 | + } |
| 152 | + }, |
| 153 | + "nbformat": 4, |
| 154 | + "nbformat_minor": 5 |
| 155 | +} |
0 commit comments