"First let's load the tools that we will need later"
]
},
{
...
...
@@ -52,7 +52,7 @@
"cell_type": "markdown",
"metadata": {},
"source": [
"Get the time series data "
"I get the time series data (stored as .csv)"
]
},
{
...
...
@@ -132,6 +132,15 @@
"name_to_plot_1"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"List of variables\n",
"\n",
"> (names are messy!) variable names need cleaning later."
]
},
{
"cell_type": "code",
"execution_count": 6,
...
...
@@ -163,16 +172,22 @@
"cell_type": "markdown",
"metadata": {},
"source": [
"Get all the sst differences, sst diff statistics and bin statistics for two data sets of different resolution."
"Here I gather:\n",
"\n",
" - the sst differences\n",
" \n",
" - sst diff statistics\n",
" \n",
" - bin statistics \n",
"\n",
"For two data sets of different resolution."
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"Now we plot everything together. \n",
"\n",
"### SST differences and the sst difference statistics across the Southern Ocean"
"### SST differences and spatial average statistics"
]
},
{
...
...
@@ -250,7 +265,7 @@
"source": [
"In both data sets we observed a pick of SST differences between 1993 to 1998. \n",
"\n",
"- Do we see the same in yearly averages and seasonal averages?"
"### - Do we see the same in yearly averages and seasonal averages?"
]
},
{
...
...
@@ -333,23 +348,27 @@
"cell_type": "markdown",
"metadata": {},
"source": [
"Large SST differences between both ESA-CCI-SST data sets persevere from 1993 - 1998. By estimating seasonal and yearly averages, the SST differences are less pronounced but they are always there. Differences are smaller in magnitude when the data has been resampled to a 1 degreee resolution and into a higher temporal resolution."
"- Large SST differences between both ESA-CCI-SST data sets persevere from 1993 - 1998.\n",
"- By estimating seasonal and yearly averages, the SST differences are less pronounced but they are always there. \n",
"- Differences are smaller in magnitude when the data has been resampled to a 1 degreee resolution and into a higher temporal resolution."
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"### SST Bin statistics for the interpolated data across the Southern Ocean"
"### SST mean Bin statistics for the interpolated data"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"Bin statistics, are the statistics recorded for each grid cell of 1 or 2 degree resolution, containing the ESA-CCI-SST data interpolated to buoy coordinates. All the interpolated points were averaged within 1 or 2 degree cells. For each cell std, median and standard error were recorded and are shown in the image below.\n",
"Bin statistics, are the statistics recorded for each grid cell of the 1 or 2 degree resolution interpolated data, that is ESA-CCI-SST data interpolated to buoy coordinates. \n",
"\n",
"All the interpolated points were averaged within 1 or 2 degree cells. For each cell **std, median and standard error** were recorded and are shown in the image below.\n",
"\n",
"Each bin statistics variable has been resampled monthly and then average over the entire S.O. "
"Each bin statistics variable has been resampled agained monthly, yearly, seasonal and then averaged over the entire S.O. Here I only show monthly outputs"
]
},
{
...
...
@@ -464,9 +483,9 @@
"cell_type": "markdown",
"metadata": {},
"source": [
"Her I read a netcdf containing the monthly means of both ESA-CCI-SST datasets. So I can compare them spatially before averaging the data over the entire Southern Ocean.\n",
"First I saved the data in another netcdf containing the monthly means of both ESA-CCI-SST datasets to visualise the data faster. Now I can compare them spatially before averaging the data over the entire Southern Ocean.\n",
"\n",
"I read the file and select only the period of 1993 to 1998. "
"I read the new file and select only the period of 1993 to 1998. "
]
},
{
...
...
@@ -481,7 +500,7 @@
},
{
"cell_type": "code",
"execution_count": 13,
"execution_count": 12,
"metadata": {},
"outputs": [],
"source": [
...
...
@@ -493,7 +512,7 @@
},
{
"cell_type": "code",
"execution_count": 14,
"execution_count": 13,
"metadata": {},
"outputs": [
{
...
...
@@ -504,7 +523,7 @@
" 'units': 'kelvin'}"
]
},
"execution_count": 14,
"execution_count": 13,
"metadata": {},
"output_type": "execute_result"
}
...
...
@@ -517,22 +536,29 @@
"cell_type": "markdown",
"metadata": {},
"source": [
"### Interactive plotting to identify big differences"
"### Interactive visualisation to identify big differences"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"Now I'll use interactive plotting to identify the region in the Southern Ocean that is causing a spike on the difference in SST.\n",
"Now I'll use interactive plotting to identify the region in the Southern Ocean that is causing a spike on the difference in SST. I builded a Bokeh App, but unfortunatly Gitlab is not the best place to share it since the size is bigger than 10 MB. Anything above that will not render into the html gitlab page.\n",
"\n",
"I builded a Bokeh App, but unfortunatly Gitlab is not the best place to share it since the size is bigguer than 10 MB.\n",
"However, you can download the [ouput.html](https://git.noc.ac.uk/brecinosrivas/orchestra-sst/-/blob/master/notebooks/output.html) to your local pc and see it in your browser.\n",
"\n",
"However you can download the [ouput.html](https://git.noc.ac.uk/brecinosrivas/orchestra-sst/-/blob/master/notebooks/output.html) to your local pc and see it in your browser.\n",
"> Make sure you down load the file, dont just open in browser. The code to build the app is below. "
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"Small animation to make my point:\n",
"\n",
"> Make sure you down load the file, dont just open in browser. \n",
"On the video we can see that in 1994/Sep there is a big difference in SST. By looking at the time series and scrolling along the maps over the entire period 1993 to 1998. I find several grid-cells at the edge of Australia and New Zeland where the difference in SST is > 200 K!"
]
},
{
...
...
@@ -619,18 +645,12 @@
"cell_type": "markdown",
"metadata": {},
"source": [
"Here is the output as a small video to make my point. \n",
"On the video we can see that in 1994/Sep there is a big difference in SST. By looking at the maps over the entire period 1993 to 1998. I find several grid-cells at the edge of Australia and New Zeland where the difference in SST is > 200 K!\n",
"\n",
"Focusing on just one spike and zooming to the region:"
"## Focusing on just one spike and zooming to the region:"
]
},
{
"cell_type": "code",
"execution_count": 15,
"execution_count": 14,
"metadata": {},
"outputs": [
{
...
...
@@ -652,7 +672,7 @@
},
{
"cell_type": "code",
"execution_count": 16,
"execution_count": 15,
"metadata": {},
"outputs": [
{
...
...
@@ -729,15 +749,10 @@
"cell_type": "markdown",
"metadata": {},
"source": [
"We can see that in the Coast of New Zeland there are some pixels in the Interpolated data set that are extremly cold compared to the coarsen data set. Is this due to stranded buoys? "
"We can see that in the Coast of New Zeland there are some pixels in the Interpolated data set that are extremly cold compared to the coarsen data set. \n",
