The following image\u00a0summarizes the answer to this question.<\/p>\n
<\/p>\n
On the left we\u00a0have\u00a0our\u00a0points.\u00a0They\u00a0are\u00a0distributed\u00a0in the\u00a0XY\u00a0space, exactly the\u00a0distribution\u00a0we\u00a0usually study\u00a0and analyse by using the\u00a0ArcGis\u00a0tools.\u00a0But\u00a0we have several points located in the same place that\u00a0correspond\u00a0to different\u00a0moments.\u00a0This is the reason why the points are also distributed along the Z axis, the time axis. The\u00a0space-time\u00a0cube\u00a0will\u00a0allow\u00a0us\u00a0to analyse\u00a0the distribution of\u00a0our\u00a0points\u00a0on\u00a0these\u00a0three\u00a0axes.<\/p>\n First<\/strong>\u00a0important<\/strong>\u00a0thing<\/strong>\u00a0to<\/strong>\u00a0remember<\/strong>: The<\/strong>\u00a0tools<\/strong>\u00a0that<\/strong>\u00a0we are<\/strong>\u00a0going<\/strong>\u00a0to be using to analyse<\/strong>\u00a0the<\/strong>\u00a0spatial-temporal<\/strong>\u00a0cube are not apt to perform the analysis of a given phenomenon based on its<\/strong>\u00a0attributes<\/strong>.\u00a0They<\/strong>\u00a0are<\/strong>\u00a0limited<\/strong>\u00a0to<\/strong>\u00a0the analysis<\/strong>\u00a0of<\/strong>\u00a0the occurrence<\/strong>\u00a0or<\/strong>\u00a0non-occurrence of a<\/strong>\u00a0phenomenon<\/strong>\u00a0(the point exists<\/strong>\u00a0or<\/strong>\u00a0does<\/strong>\u00a0not<\/strong>\u00a0exist<\/strong>).<\/strong><\/p>\n The\u00a0process\u00a0is\u00a0simple; we\u00a0will\u00a0determine\u00a0the size of a box\u00a0\u00a0(bin),\u00a0such\u00a0as pixels\u00a0or\u00a0raster cells,\u00a0but\u00a0in three dimensions.\u00a0And we\u00a0will\u00a0determine\u00a0the number\u00a0of points\u00a0contained\u00a0in\u00a0each\u00a0box.\u00a0It is\u00a0this\u00a0number\u00a0that\u00a0we\u00a0will\u00a0visualize and analyse.<\/p>\n Let\u2019s use an\u00a0example in order to follow the process in detail. \u00a0We have the weather forecast containing, that includes the gusty winds forecast. These forecasts are for a total of 8 days, every 3 hours, for an area that includes the Brittany.<\/p>\n A look at the\u00a0attribute\u00a0table\u00a0let us\u00a0see\u00a0the attribute\u00a0\u201d\u00a0gust\u00a0\u00ab\u00a0, the\u00a0speed\u00a0of the wind gusts in knots, and, another field named Datehour, indicating the date and hour\u00a0of\u00a0each\u00a0forecast.<\/p>\n Maybe\u00a0one\u00a0day it will be possible to\u00a0analyse the distribution of the\u00a0gust values\u00a0as a\u00a0function\u00a0of time.\u00a0But for the time being,\u00a0we\u00a0are\u00a0limited\u00a0to\u00a0study\u00a0the occurrence\u00a0of the points, we\u00a0can, for\u00a0example,\u00a0define\u00a0a\u00a0threshold\u00a0of\u00a0speed\u00a0that we consider as risky and analyse the distribution of the points which exceed\u00a0this\u00a0threshold\u00a0value.<\/p>\n If we\u00a0define\u00a0the\u00a0threshold\u00a0value as 20\u00a0knots\u00a0we will create\u00a0a\u00a0new\u00a0layer\u00a0of\u00a0data\u00a0containing only\u00a0the points exceeding this value. The points that do not exceed this value will not be present.\u00a0We\u00a0do\u00a0that\u00a0by\u00a0selection\u00a0of\u00a0attributes, gust> = 20,\u00a0and\u00a0we\u00a0export\u00a0the\u00a0selected\u00a0points.<\/p>\n That’s it; we\u00a0have\u00a0now\u00a0a\u00a0layer\u00a0to\u00a0manufacture\u00a0our\u00a0cube.<\/p>\n To\u00a0create\u00a0the cube, which will be\u00a0stored\u00a0under\u00a0a\u00a0NetCDF\u00a0format\u00a0(\u00a0.nc\u00a0), we\u00a0will\u00a0use\u00a0the tool\u00a0\u00a0\u00a0Create<\/em><\/strong>\u00a0a<\/em><\/strong>\u00a0space-time<\/em><\/strong>\u00a0cube<\/em><\/strong>\u00a0Toolbox\u00a0\u00a0\u00a0Tools<\/em><\/strong>\u00a0models for exploration<\/em><\/strong>\u00a0spatial-temporal<\/em><\/strong>\u00a0.<\/em><\/strong><\/p>\n Once issued\u00a0the\u00a0order,\u00a0the\u00a0settings\u00a0window\u00a0appears.<\/p>\n The\u00a0\u00a0\u00a0input<\/em>\u00a0entities<\/em>\u00a0\u00a0\u00a0correspond to\u00a0our\u00a0layer\u00a0of selected gusts\u00a0( >\u00a020\u00a0knots).<\/p>\n The\u00a0\u00a0\u00a0space-time<\/em>\u00a0cube<\/em>\u00a0output<\/em>\u00a0\u00a0\u00a0will be\u00a0our\u00a0cube.<\/p>\n The\u00a0\u00a0\u00a0time<\/em>\u00a0field<\/em>\u00a0\u00a0\u00a0will serve\u00a0to\u00a0define\u00a0the\u00a0Z\u00a0axis\u00a0and it must\u00a0absolutely\u00a0be\u00a0it\u00a0a\u00a0Date type field.\u00a0For\u00a0those\u00a0who\u00a0are\u00a0always\u00a0with\u00a0shapefiles,\u00a0remember\u00a0that\u00a0the date fields can\u00a0only\u00a0contain\u00a0the date\u00a0but never the time.\u00a0If\u00a0you\u00a0are going\u00a0to work\u00a0with time lapses of less than 24 hours (a day) you\u00a0must\u00a0have\u00a0a class\u00a0of geodatabase entities\u00a0mandatorily\u00a0.<\/p>\n For\u00a0the time being\u00a0we\u00a0do\u00a0not\u00a0have count with a\u00a0model\u00a0cube\u00a0to recover\u00a0the\u00a0others\u00a0cubes information.\u00a0If we\u00a0make\u00a0a\u00a0series\u00a0of cubes, and the first serves \u00a0as\u00a0model\u00a0, we can be certain that\u00a0the steps\u00a0and\u00a0the location of the\u00a0boxes\u00a0will be the\u00a0same\u00a0for\u00a0all\u00a0the cubes\u00a0produced\u00a0.<\/p>\n The\u00a0time interval<\/em>\u00a0\u00a0is\u00a0the\u00a0size\u00a0of\u00a0the cube\u00a0boxes\u00a0on\u00a0the\u00a0Z\u00a0axis\u00a0. In\u00a0others\u00a0terms, the time step\u00a0that\u00a0we are\u00a0going\u00a0to apply\u00a0to\u00a0our\u00a0analysis.\u00a0In\u00a0our example\u00a0, the time step of the\u00a0forecasts\u00a0is\u00a03\u00a0hours\u00a0.\u00a0A\u00a0size\u00a012\u00a0hour\u00a0box\u00a0will produce 16\u00a0 time intervals in Z.\u00a0A box\u00a0size\u00a0of 24\u00a0hours,\u00a0will\u00a0produce 8.<\/p>\n One\u00a0of the\u00a0constraints\u00a0of the cube\u00a0is\u00a0it\u00a0must contain, at least 10 time lapses.\u00a0\u00a0Therefore we\u00a0will choose\u00a0so\u00a0a time lapse of 12\u00a0hours\u00a0.<\/p>\n The\u00a0distance\u00a0interval<\/em>\u00a0\u00a0\u00a0is\u00a0the\u00a0cell\u00a0size\u00a0in\u00a0the XY plane.<\/p>\n If\u00a0you\u00a0do\u00a0not\u00a0select one\u00a0or\u00a0both\u00a0intervals, the program\u00a0will calculate\u00a0one by\u00a0default\u00a0.<\/p>\n To visualize the cube\u00a0you\u00a0must\u00a0Download\u00a0a\u00a0Toolkit\u00a0containing\u00a0two\u00a0tools:\u00a0one\u00a0that\u00a0allows\u00a0to visualize the cube in 2D and\u00a0another\u00a0that\u00a0allows\u00a0to visualize it in 3D. Once downloaded\u00a0and\u00a0decompressed,\u00a0you\u00a0can\u00a0add\u00a0it in\u00a0the\u00a0geoprocessor\u00a0by\u00a0clicking\u00a0on\u00a0the\u00a0ribbon\u00a0Create\u00a0->\u00a0Toolbox<\/p>\n We\u00a0will\u00a0see\u00a0the visualization in 3D.<\/p>\n You\u00a0must\u00a0have\u00a0a\u00a03D\u00a0window\u00a0(Scene)\u00a0in\u00a0your\u00a0project\u00a0.\u00a0If\u00a0you\u00a0do not, please make sure to,\u00a0click\u00a0\u00a0Insert\u00a0-> New Scene You\u00a0must\u00a0remove\u00a0the\u00a0reference\u00a0to\u00a0a\u00a0layer\u00a0of elevation.\u00a0If you do not, the\u00a0columns\u00a0of the cube\u00a0will be\u00a0displayed\u00a0relative to\u00a0the\u00a0ground\u00a0level\u00a0and you will\u00a0loose\u00a0visibility\u00a0of the\u00a0correspondence\u00a0of\u00a0the\u00a0different\u00a0time lapses.<\/p>\n In order to do this,\u00a0go to panel\u00a0Content,\u00a0click\u00a0on your\u00a0scene ->\u00a0properties\u00a0-> Elevation surface,\u00a0deploy\u00a0Ground\u00a0and\u00a0delete\u00a0all\u00a0services\u00a0present.<\/p>\n Execute\u00a0the\u00a0visualization\u00a0command\u00a0of the\u00a0spatial-temporal\u00a0cube\u00a0in 3D:<\/p>\n <\/p>\n![\"\"](\"https:\/\/i0.wp.com\/www.sigterritoires.fr\/wp-content\/uploads\/2018\/03\/cubo-1.png\")
<\/a><\/p>\n<\/a><\/p>\n<\/a><\/p>\nCreation<\/strong>\u00a0of the<\/strong>\u00a0space-time<\/strong>\u00a0cube<\/strong><\/h3>\n
<\/a><\/p>\n<\/a><\/p>\nVisualization of the<\/strong>\u00a0space-time<\/strong>\u00a0cube<\/strong><\/h3>\n
\nThe\u00a0Toolbox\u00a0East\u00a0downloadable\u00a0at\u00a0http:\/\/esriurl.com\/SpaceTimeCubeUtilities<\/a>.<\/p>\n<\/a><\/p>\n
\n<\/a><\/p>\n