#
`binind2latlon` documentation

`binind2latlon` converts binned index values of a sinusoidal grid to geocoordinates.

More information on the sinusoidal grid here.

## Contents

## Syntax

[lat,lon] = binind2latlon(binind) [lat,lon] = binind2latlon(...,'rows',NumberOfRows)

## Description

`[lat,lon] = binind2latlon(binind)` gives the geo coordinates for the bin indices `binind`.

`[lat,lon] = binind2latlon(...,'rows',NumberOfRows)` specifies the number of rows in the grid. By default, binind2latlon will try to figure out the number of rows automatically, but there's no guarantee it'll work. If you know the number of rows, specify them just to be sure. According to the Resolutions table here, common numbers of rows are as follows:

- 180 rows for 1 a degree grid
- 2160 rows for 5 minute grid
- 4320 rows for a 1.5 minute grid

## Example 1: A small grid

Let's recreate the simple 18 row example seen here. Start by plotting national borders with my `borders` function:

borders('countries','k-','nomappingtoolbox') axis tight xlabel 'longitude' ylabel 'latitude'

The example on the NASA site contains 412 bins that span 18 rows of latitude:

% List of bin numbers: bins = (1:412)'; % What geocoordinates correspond to each bin number? [lat,lon] = binind2latlon(bins,'rows',18); % Label each bin centered on its geocoordinate: text(lon,lat,num2str(bins),'color','r','horiz','center','fontsize',6)

## Example 2: Actual data

This example uses some monthly CHL data found here. Start by reading in the data. Although these files end in .nc, the `ncread` function does not recognize the datatype for the variables we want to plot. So we have to use `h5read` instead:

A = h5read('A20021822002212.L3b_MO_CHL.nc','/level-3_binned_data/chlor_a'); B = h5read('A20021822002212.L3b_MO_CHL.nc','/level-3_binned_data/BinList'); % For convenience, assign z and bins as variables: z = double(A.sum_squared); bins = B.bin_num;

The geocoordinates of each of those `z` values can be found quite easily:

[lat,lon] = binind2latlon(bins);

Notice there are over 11 million measurements here, so we might want to look at just the measurements in a small area, say the Indian Ocean, in the region (35E-110E, 40S-20N). Get the indices corresponding to these coordinates:

ind = lon>=35 & lon<=110 & lat>=-40 &lat<=20;

And unfortunately they don't make a perfect grid, so if you want gridded data you'll have to use `griddata` or `scatteredInterpolant`. Or you can make a scatterplot of the data. Old versions of Matlab like the one I'm using choke when you give `scatter` a lot of data points, so below I'm using Aslak Grinsted's `fastscatter` function. I'm also using the `cmocean` (Thyng et al., 2016) colormap.

figure fastscatter(lon(ind),lat(ind),z(ind),'markersize',.1) axis tight caxis([0 4]) cb = colorbar; ylabel(cb,'sum squared') cmocean algae borders('countries','facecolor',0.6*[1 1 1],'nomappingtoolbox')

## Author Info

This function and supporting documentation were written by Chad A. Greene of the University of Texas at Austin.