|
| 1 | +{ |
| 2 | + "cells": [ |
| 3 | + { |
| 4 | + "cell_type": "markdown", |
| 5 | + "metadata": {}, |
| 6 | + "source": [ |
| 7 | + "# Stommel Gyre on Unstructured Grid\n", |
| 8 | + "This tutorial walks through creating a UXArray dataset using the Stommel Gyre analytical solution for a closed rectangular domain on a beta-plane" |
| 9 | + ] |
| 10 | + }, |
| 11 | + { |
| 12 | + "cell_type": "code", |
| 13 | + "execution_count": null, |
| 14 | + "metadata": {}, |
| 15 | + "outputs": [], |
| 16 | + "source": [ |
| 17 | + "def stommel_fieldset_uxarray(xdim=200, ydim=200):\n", |
| 18 | + " \"\"\"Simulate a periodic current along a western boundary, with significantly\n", |
| 19 | + " larger velocities along the western edge than the rest of the region\n", |
| 20 | + "\n", |
| 21 | + " The original test description can be found in: N. Fabbroni, 2009,\n", |
| 22 | + " Numerical Simulation of Passive tracers dispersion in the sea,\n", |
| 23 | + " Ph.D. dissertation, University of Bologna\n", |
| 24 | + " http://amsdottorato.unibo.it/1733/1/Fabbroni_Nicoletta_Tesi.pdf\n", |
| 25 | + " \"\"\"\n", |
| 26 | + " import math\n", |
| 27 | + "\n", |
| 28 | + " import numpy as np\n", |
| 29 | + " import pandas as pd\n", |
| 30 | + " import uxarray as ux\n", |
| 31 | + "\n", |
| 32 | + " a = b = 66666 * 1e3\n", |
| 33 | + " scalefac = 0.00025 # to scale for physically meaningful velocities\n", |
| 34 | + "\n", |
| 35 | + " # Coordinates of the test fieldset\n", |
| 36 | + " # Crowd points to the west edge of the domain\n", |
| 37 | + " # using a polyonmial map on x-direction\n", |
| 38 | + " x = np.linspace(0, 1, xdim, dtype=np.float32)\n", |
| 39 | + " lon, lat = np.meshgrid(a * x, np.linspace(0, b, ydim, dtype=np.float32))\n", |
| 40 | + " points = (lon.flatten() / 1111111.111111111, lat.flatten() / 1111111.111111111)\n", |
| 41 | + "\n", |
| 42 | + " # Create the grid\n", |
| 43 | + " uxgrid = ux.Grid.from_points(points, method=\"regional_delaunay\")\n", |
| 44 | + " uxgrid.construct_face_centers()\n", |
| 45 | + "\n", |
| 46 | + " # Define arrays U (zonal), V (meridional) and P (sea surface height)\n", |
| 47 | + " U = np.zeros((1, 1, lat.size), dtype=np.float32)\n", |
| 48 | + " V = np.zeros((1, 1, lat.size), dtype=np.float32)\n", |
| 49 | + " P = np.zeros((1, 1, lat.size), dtype=np.float32)\n", |
| 50 | + "\n", |
| 51 | + " beta = 2e-11\n", |
| 52 | + " r = 1 / (11.6 * 86400)\n", |
| 53 | + " es = r / (beta * a)\n", |
| 54 | + "\n", |
| 55 | + " i = 0\n", |
| 56 | + " for x, y in zip(lon.flatten(), lat.flatten()):\n", |
| 57 | + " xi = x / a\n", |
| 58 | + " yi = y / b\n", |
| 59 | + " P[0, 0, i] = (\n", |
| 60 | + " (1 - math.exp(-xi / es) - xi) * math.pi * np.sin(math.pi * yi) * scalefac\n", |
| 61 | + " )\n", |
| 62 | + " U[0, 0, i] = (\n", |
| 63 | + " -(1 - math.exp(-xi / es) - xi)\n", |
| 64 | + " * math.pi**2\n", |
| 65 | + " * np.cos(math.pi * yi)\n", |
| 66 | + " * scalefac\n", |
| 67 | + " )\n", |
| 68 | + " V[0, 0, i] = (\n", |
| 69 | + " (math.exp(-xi / es) / es - 1) * math.pi * np.sin(math.pi * yi) * scalefac\n", |
| 70 | + " )\n", |
| 71 | + " i += 1\n", |
| 72 | + "\n", |
| 73 | + " u = ux.UxDataArray(\n", |
| 74 | + " data=U,\n", |
| 75 | + " name=\"u\",\n", |
| 76 | + " uxgrid=uxgrid,\n", |
| 77 | + " dims=[\"time\", \"nz1\", \"n_node\"],\n", |
| 78 | + " coords=dict(\n", |
| 79 | + " time=([\"time\"], pd.to_datetime([\"2000-01-01\"])),\n", |
| 80 | + " nz1=([\"nz1\"], [0]),\n", |
| 81 | + " ),\n", |
| 82 | + " attrs=dict(\n", |
| 83 | + " description=\"zonal velocity\",\n", |
| 84 | + " units=\"m/s\",\n", |
| 85 | + " location=\"node\",\n", |
| 86 | + " mesh=\"delaunay\",\n", |
| 87 | + " ),\n", |
| 88 | + " )\n", |
| 89 | + " v = ux.UxDataArray(\n", |
| 90 | + " data=V,\n", |
| 91 | + " name=\"v\",\n", |
| 92 | + " uxgrid=uxgrid,\n", |
| 93 | + " dims=[\"time\", \"nz1\", \"n_node\"],\n", |
| 94 | + " coords=dict(\n", |
| 95 | + " time=([\"time\"], pd.to_datetime([\"2000-01-01\"])),\n", |
| 96 | + " nz1=([\"nz1\"], [0]),\n", |
| 97 | + " ),\n", |
| 98 | + " attrs=dict(\n", |
| 99 | + " description=\"meridional velocity\",\n", |
| 100 | + " units=\"m/s\",\n", |
| 101 | + " location=\"node\",\n", |
| 102 | + " mesh=\"delaunay\",\n", |
| 103 | + " ),\n", |
| 104 | + " )\n", |
| 105 | + " p = ux.UxDataArray(\n", |
| 106 | + " data=P,\n", |
| 107 | + " name=\"p\",\n", |
| 108 | + " uxgrid=uxgrid,\n", |
| 109 | + " dims=[\"time\", \"nz1\", \"n_node\"],\n", |
| 110 | + " coords=dict(\n", |
| 111 | + " time=([\"time\"], pd.to_datetime([\"2000-01-01\"])),\n", |
| 112 | + " nz1=([\"nz1\"], [0]),\n", |
| 113 | + " ),\n", |
| 114 | + " attrs=dict(\n", |
| 115 | + " description=\"pressure\",\n", |
| 116 | + " units=\"N/m^2\",\n", |
| 117 | + " location=\"node\",\n", |
| 118 | + " mesh=\"delaunay\",\n", |
| 119 | + " ),\n", |
| 120 | + " )\n", |
| 121 | + "\n", |
| 122 | + " return ux.UxDataset({\"u\": u, \"v\": v, \"p\": p}, uxgrid=uxgrid)\n", |
| 123 | + "\n", |
| 124 | + "\n", |
| 125 | + "uxds = stommel_fieldset_uxarray(50, 50)\n", |
| 126 | + "\n", |
| 127 | + "uxds.uxgrid.plot(\n", |
| 128 | + " line_width=0.5,\n", |
| 129 | + " height=500,\n", |
| 130 | + " width=1000,\n", |
| 131 | + " title=\"Regional Delaunay Regions\",\n", |
| 132 | + ")" |
| 133 | + ] |
| 134 | + }, |
| 135 | + { |
| 136 | + "cell_type": "code", |
| 137 | + "execution_count": null, |
| 138 | + "metadata": {}, |
| 139 | + "outputs": [], |
| 140 | + "source": [ |
| 141 | + "def stommel_fieldset_xarray(xdim=200, ydim=200, grid_type=\"A\"):\n", |
| 142 | + " \"\"\"Simulate a periodic current along a western boundary, with significantly\n", |
| 143 | + " larger velocities along the western edge than the rest of the region\n", |
| 144 | + "\n", |
| 145 | + " The original test description can be found in: N. Fabbroni, 2009,\n", |
| 146 | + " Numerical Simulation of Passive tracers dispersion in the sea,\n", |
| 147 | + " Ph.D. dissertation, University of Bologna\n", |
| 148 | + " http://amsdottorato.unibo.it/1733/1/Fabbroni_Nicoletta_Tesi.pdf\n", |
| 149 | + " \"\"\"\n", |
| 150 | + " import math\n", |
| 151 | + "\n", |
| 152 | + " import numpy as np\n", |
| 153 | + " import pandas as pd\n", |
| 154 | + " import xarray as xr\n", |
| 155 | + "\n", |
| 156 | + " a = b = 10000 * 1e3\n", |
| 157 | + " scalefac = 0.05 # to scale for physically meaningful velocities\n", |
| 158 | + " dx, dy = a / xdim, b / ydim\n", |
| 159 | + "\n", |
| 160 | + " # Coordinates of the test fieldset (on A-grid in deg)\n", |
| 161 | + " lon = np.linspace(0, a, xdim, dtype=np.float32)\n", |
| 162 | + " lat = np.linspace(0, b, ydim, dtype=np.float32)\n", |
| 163 | + "\n", |
| 164 | + " # Define arrays U (zonal), V (meridional) and P (sea surface height)\n", |
| 165 | + " U = np.zeros((1, 1, lat.size, lon.size), dtype=np.float32)\n", |
| 166 | + " V = np.zeros((1, 1, lat.size, lon.size), dtype=np.float32)\n", |
| 167 | + " P = np.zeros((1, 1, lat.size, lon.size), dtype=np.float32)\n", |
| 168 | + "\n", |
| 169 | + " beta = 2e-11\n", |
| 170 | + " r = 1 / (11.6 * 86400)\n", |
| 171 | + " es = r / (beta * a)\n", |
| 172 | + "\n", |
| 173 | + " for j in range(lat.size):\n", |
| 174 | + " for i in range(lon.size):\n", |
| 175 | + " xi = lon[i] / a\n", |
| 176 | + " yi = lat[j] / b\n", |
| 177 | + " P[..., j, i] = (\n", |
| 178 | + " (1 - math.exp(-xi / es) - xi)\n", |
| 179 | + " * math.pi\n", |
| 180 | + " * np.sin(math.pi * yi)\n", |
| 181 | + " * scalefac\n", |
| 182 | + " )\n", |
| 183 | + " if grid_type == \"A\":\n", |
| 184 | + " U[..., j, i] = (\n", |
| 185 | + " -(1 - math.exp(-xi / es) - xi)\n", |
| 186 | + " * math.pi**2\n", |
| 187 | + " * np.cos(math.pi * yi)\n", |
| 188 | + " * scalefac\n", |
| 189 | + " )\n", |
| 190 | + " V[..., j, i] = (\n", |
| 191 | + " (math.exp(-xi / es) / es - 1)\n", |
| 192 | + " * math.pi\n", |
| 193 | + " * np.sin(math.pi * yi)\n", |
| 194 | + " * scalefac\n", |
| 195 | + " )\n", |
| 196 | + "\n", |
| 197 | + " time = pd.to_datetime([\"2000-01-01\"])\n", |
| 198 | + " z = [0]\n", |
| 199 | + " if grid_type == \"C\":\n", |
| 200 | + " V[..., :, 1:] = (P[..., :, 1:] - P[..., :, 0:-1]) / dx * a\n", |
| 201 | + " U[..., 1:, :] = -(P[..., 1:, :] - P[..., 0:-1, :]) / dy * b\n", |
| 202 | + " u_dims = [\"time\", \"nz1\", \"face_lat\", \"node_lon\"]\n", |
| 203 | + " u_lat = lat\n", |
| 204 | + " u_lon = lon - dx * 0.5\n", |
| 205 | + " u_location = \"x_edge\"\n", |
| 206 | + " v_dims = [\"time\", \"nz1\", \"node_lat\", \"face_lon\"]\n", |
| 207 | + " v_lat = lat - dy * 0.5\n", |
| 208 | + " v_lon = lon\n", |
| 209 | + " v_location = \"y_edge\"\n", |
| 210 | + " p_dims = [\"time\", \"nz1\", \"face_lat\", \"face_lon\"]\n", |
| 211 | + " p_lat = lat\n", |
| 212 | + " p_lon = lon\n", |
| 213 | + " p_location = \"face\"\n", |
| 214 | + "\n", |
| 215 | + " else:\n", |
| 216 | + " u_dims = [\"time\", \"nz1\", \"node_lat\", \"node_lon\"]\n", |
| 217 | + " v_dims = [\"time\", \"nz1\", \"node_lat\", \"node_lon\"]\n", |
| 218 | + " p_dims = [\"time\", \"nz1\", \"node_lat\", \"node_lon\"]\n", |
| 219 | + " u_lat = lat\n", |
| 220 | + " u_lon = lon\n", |
| 221 | + " v_lat = lat\n", |
| 222 | + " v_lon = lon\n", |
| 223 | + " u_location = \"node\"\n", |
| 224 | + " v_location = \"node\"\n", |
| 225 | + " p_lat = lat\n", |
| 226 | + " p_lon = lon\n", |
| 227 | + " p_location = \"node\"\n", |
| 228 | + "\n", |
| 229 | + " u = xr.DataArray(\n", |
| 230 | + " data=U,\n", |
| 231 | + " name=\"u\",\n", |
| 232 | + " dims=u_dims,\n", |
| 233 | + " coords=[time, z, u_lat, u_lon],\n", |
| 234 | + " attrs=dict(\n", |
| 235 | + " description=\"zonal velocity\",\n", |
| 236 | + " units=\"m/s\",\n", |
| 237 | + " location=u_location,\n", |
| 238 | + " mesh=f\"Arakawa-{grid_type}\",\n", |
| 239 | + " ),\n", |
| 240 | + " )\n", |
| 241 | + " v = xr.DataArray(\n", |
| 242 | + " data=V,\n", |
| 243 | + " name=\"v\",\n", |
| 244 | + " dims=v_dims,\n", |
| 245 | + " coords=[time, z, v_lat, v_lon],\n", |
| 246 | + " attrs=dict(\n", |
| 247 | + " description=\"meridional velocity\",\n", |
| 248 | + " units=\"m/s\",\n", |
| 249 | + " location=v_location,\n", |
| 250 | + " mesh=f\"Arakawa-{grid_type}\",\n", |
| 251 | + " ),\n", |
| 252 | + " )\n", |
| 253 | + " p = xr.DataArray(\n", |
| 254 | + " data=P,\n", |
| 255 | + " name=\"p\",\n", |
| 256 | + " dims=p_dims,\n", |
| 257 | + " coords=[time, z, p_lat, p_lon],\n", |
| 258 | + " attrs=dict(\n", |
| 259 | + " description=\"pressure\",\n", |
| 260 | + " units=\"N/m^2\",\n", |
| 261 | + " location=p_location,\n", |
| 262 | + " mesh=f\"Arakawa-{grid_type}\",\n", |
| 263 | + " ),\n", |
| 264 | + " )\n", |
| 265 | + "\n", |
| 266 | + " return xr.Dataset({\"u\": u, \"v\": v, \"p\": p})\n", |
| 267 | + "\n", |
| 268 | + "\n", |
| 269 | + "ds_arakawa_a = stommel_fieldset_xarray(50, 50, \"A\")\n", |
| 270 | + "ds_arakawa_c = stommel_fieldset_xarray(50, 50, \"C\")" |
| 271 | + ] |
| 272 | + }, |
| 273 | + { |
| 274 | + "cell_type": "code", |
| 275 | + "execution_count": null, |
| 276 | + "metadata": {}, |
| 277 | + "outputs": [], |
| 278 | + "source": [ |
| 279 | + "ds_arakawa_a" |
| 280 | + ] |
| 281 | + }, |
| 282 | + { |
| 283 | + "cell_type": "code", |
| 284 | + "execution_count": null, |
| 285 | + "metadata": {}, |
| 286 | + "outputs": [], |
| 287 | + "source": [ |
| 288 | + "ds_arakawa_a[\"u\"].attrs" |
| 289 | + ] |
| 290 | + }, |
| 291 | + { |
| 292 | + "cell_type": "code", |
| 293 | + "execution_count": null, |
| 294 | + "metadata": {}, |
| 295 | + "outputs": [], |
| 296 | + "source": [ |
| 297 | + "ds_arakawa_c" |
| 298 | + ] |
| 299 | + }, |
| 300 | + { |
| 301 | + "cell_type": "code", |
| 302 | + "execution_count": null, |
| 303 | + "metadata": {}, |
| 304 | + "outputs": [], |
| 305 | + "source": [ |
| 306 | + "import numpy as np\n", |
| 307 | + "\n", |
| 308 | + "min_length_scale = 1111111.111111111 * np.sqrt(np.min(uxds.uxgrid.face_areas))\n", |
| 309 | + "print(min_length_scale)\n", |
| 310 | + "\n", |
| 311 | + "max_v = np.sqrt(uxds[\"u\"] ** 2 + uxds[\"v\"] ** 2).max()\n", |
| 312 | + "print(max_v)\n", |
| 313 | + "\n", |
| 314 | + "cfl = 0.1\n", |
| 315 | + "dt = cfl * min_length_scale / max_v\n", |
| 316 | + "print(dt)" |
| 317 | + ] |
| 318 | + }, |
| 319 | + { |
| 320 | + "cell_type": "code", |
| 321 | + "execution_count": null, |
| 322 | + "metadata": {}, |
| 323 | + "outputs": [], |
| 324 | + "source": [ |
| 325 | + "from datetime import timedelta\n", |
| 326 | + "\n", |
| 327 | + "import numpy as np\n", |
| 328 | + "import uxarray as ux\n", |
| 329 | + "\n", |
| 330 | + "from parcels import Particle, ParticleSet, UxAdvectionEuler, UXFieldSet\n", |
| 331 | + "\n", |
| 332 | + "npart = 10\n", |
| 333 | + "fieldset = UXFieldSet(uxds)\n", |
| 334 | + "# pset = ParticleSet(\n", |
| 335 | + "# fieldset,\n", |
| 336 | + "# pclass=Particle,\n", |
| 337 | + "# lon=np.linspace(1, 59, npart),\n", |
| 338 | + "# lat=np.zeros(npart)+30)\n", |
| 339 | + "# pset.execute(UxAdvectionEuler, runtime=timedelta(hours=24), dt=timedelta(seconds=dt))" |
| 340 | + ] |
| 341 | + } |
| 342 | + ], |
| 343 | + "metadata": { |
| 344 | + "kernelspec": { |
| 345 | + "display_name": "parcels", |
| 346 | + "language": "python", |
| 347 | + "name": "python3" |
| 348 | + }, |
| 349 | + "language_info": { |
| 350 | + "codemirror_mode": { |
| 351 | + "name": "ipython", |
| 352 | + "version": 3 |
| 353 | + }, |
| 354 | + "file_extension": ".py", |
| 355 | + "mimetype": "text/x-python", |
| 356 | + "name": "python", |
| 357 | + "nbconvert_exporter": "python", |
| 358 | + "pygments_lexer": "ipython3", |
| 359 | + "version": "3.13.2" |
| 360 | + } |
| 361 | + }, |
| 362 | + "nbformat": 4, |
| 363 | + "nbformat_minor": 2 |
| 364 | +} |
0 commit comments