The Top Data Sources for World Land Cover
Over the span of your life, how much has the Earth changed? Quite a bit, but you probably just can’t put a number to it.
Related to magnetic card reader model zcs100 ic software download, The nationwide debt counter, and that’s also called the countrywide debt clock, is regarded as a display about the size of the billboard bearing a resemblance to your digital clock that keeps a jogging total that may be constantly updating the present state belonging to the U.s. Gross national financial debt at the same time.
Thanks to global land cover and satellite sensors like MODIS, AVHRR and Enhanced Thematic Mapper, we can finally shed some light on our changing planet.
But what global land cover data sets exists? What are the best?
What is their accuracy? How fine is its spatial resolution?
Seeing is Believing
To help you follow along and get a visual understanding of the accuracy of the land cover classification, we produce screenshots of New York. Some of the key features to keep in mind are:
CENTRAL PARK: The square green space in Manhattan, the Jacqueline Kennedy Onassis Reservoir and the Upper Bay that divides New York and New Jersey.
FIRE ISLAND: The thin barrier island off the southern shore which protects the beach fronts of Long Island, New York.
LONG ISLAND: The islands in the north-east corner of Long Island like Plum Island, Great Gull Island and Little Gull Island.
Here is New York in the eyes of Sentinel-2 pinpointing the features above.
How well are these features classified in the land cover maps? It’s not an easy task to classify the globe.
Read below for a summary of the various land cover classification schemes available for free at your finger tips.
1 Global Land Survey (GLS)
At 30-meter resolution, this land cover is one of the finest available. The University of Maryland teamed up with the USGS to lace together its circa 2010 tree cover, bare ground and persistent surface water.
Using Landsat 7 ETM+ data, its most impressive attribute is its tree cover canopy described as percentage per output grid cell. This canopy cover is used to weigh in global forest extent, loss, and gain from 2000 to 2012 such as this Global Forest Change webmap.
Studies have shown that the overall accuracy of the Global Land Survey (GLS) static forest cover to be 91% with forest cover change at >88%. Echoing the accuracy in this paper, Central Park and outer islands are clearly visible. The lack of forested area in Fire island is consistent and logical. Things are looking good.
READ MORE:4 Global Forest Maps to See the Forest for the Trees
2 Climate Change Initiative (CCI) Land Cover V2
At 300 meter resolution, ENVISAT MERIS sensor is the biggest contributor to the 3 epoch land cover maps (1998-02, 2003-07 and 2008-12) of the CCI Land Cover V2.
By area proportions, it’s been cited to reach 73% accuracy for the 23-class land cover map. Additionally, the European Space Agency has created the ESA/CCI Land Cover viewer to dynamically view the land cover
Visually, you can see how it captures the thin barrier on the southern shore of long island. Despite missing most of central park (4km x 0.8km), it classifies the north-east islands impeccably.
3 OSM Land Use Data
With all the other land cover classification, it’s based on image classification algorithms. Imagine a single composer writing a symphony classifying hundreds of satellite images into a land cover masterpiece.
But what happens when you have thousands of artists writing their own music? When each piece synchronizes from the ginormous global community known as OpenStreetMap, you get one ridiculously accurate land use map.
In a single pixel ENVISAT MERIS pixel, count them – 157 buildings
On top of that, it tells you if it’s residential, commercial, industrial, or any other type.
The only downsides to OSM land use data are:
- There are tons of data gaps.
- It probably won’t capture say a deciduous tree from a conifer.
- You have random people updating it. (with this said quality is quite good)
But in New York, OSM Land Cover reigns supreme. The outline of Central Park is picturesque. The southern shore and north-east islands are vectorized outlines. Pan over north a couple of times and the land cover is empty.
READ MORE:10 Free GIS Data Sources: Best Global Raster and Vector Datasets
4 MCD12Q1 0.5 km MODIS-based Global Land Cover Climatology
The 500-meter MODIS Land Cover Maps (17 land cover classes) describes the dominant class based on a 10 year span (2001-2010).
Studies have shown interannual variability with 40% of pixels showing change in class one or more times in the 10-year span. Because of its coarseness, it misses the mark on Central Park and Fire Island. But it captures the islands quite well.
However for all intents and purposes, it serves its job for weather and climate models.
5 USGS – Global Land Cover Characterization (GLCC)
GLCC is based on one-year Advanced Very High Resolution Radiometer (AVHRR) using an unsupervised image classification approach. Based on land area occupied, GLCC reaches a 66.9% accuracy. When the observer can’t deduce a pixel as a “true” cover, this majority rule accuracy has an accuracy even higher at 78.7℅ throwing out those sites.
GLCC is being used in a range of environmental modeling applications including the Goddard Earth Observing System Model V5 (GEOS-5).
This 1-kilometer pixel size land cover classification has classified Central Park. However, it misses the boat for Fire Island and Plum Island.
6 GlobeLand30
In hopes of resurrecting the 2000 and 2010 30-meter land cover from the National Geomatics Center of China, it’s gone missing in action.
These 30-meter resolution land cover maps show global distribution of 10 major land cover classes: water bodies, wetland, artificial surfaces, cultivated land, permanent snow and ice, forests, grasslands, shrubland, bare land and tundra.
It took over 10,000 Landsat satellite images to cover the entire Earth at 30 m resolution. This land cover uses pixel- and object-based methods and each class is identified in a prioritized sequence. At 8 selected areas, it achieves an overall classification accuracy of 80%.
We’re empty-handed for New York at present time, but the data we have for other parts of the going looks fairly reasonable. To say the least, it’s disappearance is mysterious. Nowhere to be found at the present time, we’ll update the article when or if it’s back online
READ MORE:USGS Earth Explorer: Download Free Landsat Imagery
7 UN FAO Global Land Cover Network (GLC-SHARE)
The focus for the FAO’s GLC SHARE land cover is land management. This is reflected particularly in some of its classes – cropland, grassland, bare soil and mangroves. It includes artificial surfaces, water bodies, snow, treed, shrubs, herbaceous and sparse vegetation.
It’s coarse at today’s standards being a bit sharper than 1km grid cells. Further, it has an accuracy of about 80% with 1087 validation sites.
You can see how this classification completely misses Central Park, the majority of Fire Island and north-east islands. This is partly due to its coarse resolution.
Its primary use is for better understanding land management so we’d expect better results in agricultural areas.
8 Land Cover Type Yearly L3 Global 0.05Deg CMG
The Climate Modeling Grid has the same roots as MCD12Q1 using the same supervised classification-tree algorithm. This data set is available to download from the USGS Earth Explorer.
Although it generates the same 17-classes defined by the International Geosphere Biosphere Programme, it’s at a much lower spatial resolution (0.05°).
At 5.5km pixels you can’t get the precision level compared to all the other land cover options. For the average person, it’s hard to even recognize we’re looking at New York anymore.
9 Terrapop
Terrapop contains a variety of existing global land cover data sets from the already mentioned 23-class ESA GLC and MODIS 1km.
On top of that, it includes a very coarse agricultural lands classification circa 2000 at 10km derived from the Global Landscapes Initiative. This contains harvested area and yield of 175 crops to better understand agricultural supply and demand.
We give a tutorial in great detail how to extract data from the intuitive Terrapop interface.
READ MORE:How to Get Harmonized Environmental & Demographic Data with TerraPop
Which land cover do you use?
Space agencies around the globe are launching satellites to meet the demand of accurate land cover of the planet.
Our simple visual assessment of New York shows how much land cover varies across providers.
It’s only at finer scale land cover can capture significant human activities on the land. Despite its significant gaps in data, there’s no land use as detailed as OSM.
Almost like tunnel vision, it’s surprising that scientists don’t use a multi-scale, multi-data iterative approach using the best available at a given location.
How to Make Choropleth Maps Using Data Classification
You have your quantitative data ready to go. Your mouse is hovering over “classify” waiting to generate the many colors of choropleth maps.
But you can’t help wonder if you’re choosing the right data classification mode.
Equal intervals, quantile, natural breaks, pretty breaks – there’s a lot to choose from. But what is the difference between each of them?
This post will help you understand the types of choropleth maps that exist and which one to choose for your maps.
Choose Your Number of Classes
First, you must aggregate data based on a number of classes. When you have more classes, you get more variation sometimes making it harder to separate shading. If you want to test out different shading, ColorBrewer has a tool for color advice.
For example, here’s 10 classes:
While fewer classes provides less separation between classes such as 5 classes below.
After all, the number of classes you decide with really depends on the purpose of your map.
Select Your Data Classification Method
Second, you will have to decide how to classify your data. To put it another way, data classification arranges your data with boundaries to separate classes. You could separate your classes with an equal interval mode:
Alternatively, you could select a quantile type of classifier where it arranges the data differently (more on this below)
Each data classification technique produces unique choropleth maps. But they all paint a different story to the map reader. The one thing you must realize is that you’re using the same data in each choropleth map, but what’s really changing is how you classify the data.
Our Example Data
The most important thing you have to realize is that for each of these choropleth maps we create, we use the same data. What’s changing is how we classify the data.
In this example, we count the number of letters in country names. For example:
- Mali, Cuba and Peru and others are four letter countries.
- Whereas, Bosnia and Herzegovina has 22 characters.
If you plot out 4 to 22 characters, it will have a lot of colors.
For example, the four-letter countries are the lightest shades of green. As the letter count increases, the shading gets darker.
Which country belongs to which group? It’s hard to tell.
So this is why we use data classification. When we group by classes, there’s less shading and we aggregate the data by group.
Ultimately, the question is how do we define those class boundaries or bins? In other words, how do we classify the data into groups?
First, let’s try dividing classes into evenly-spaced groupings like equal intervals below and see what happens.
Equal Interval Data Classification
Equal interval is cut and dry. All it really does is divide the classes into equal groups.
- Class 1: 4 – 8 (113 countries have four, five, six, seven or eight letters)
- Class 2: 8 – 12 (41)
- Class 3: 12 – 16 (12)
- Class 4: 16 – 20 (8)
- Class 5: 20 – 24 (2)
The minimum number of characters of a country is 4 such as Peru. The maximum number of characters is 24, which is Central African Republic. When you plot each country and their number of characters on a map, it looks like this (the brackets indicate the count):
Equal interval data classification subtracts the maximum value from minimum value (24-4=20). In our example, we generated 5 classes but the number of classes is entirely up to you. Then, it divides 20 by 5 and you get an interval (20/5=4).
Almost always, equal interval choropleth maps result in an unequal count of countries per class. For example, class 1 has 113 countries out of 176 countries with four, five, six and seven letters.
However, only 2 countries have more than 20 letters. As a result, this map displays more light shaded colors compared to only 2 with the dark shading.
But what happens if you want the count of countries in each class to be close to equal? That’s when you should use a quantile map.
Quantile (Equal Count) Classification
The quantile map tries to bin the same count of features in each of the 5 classes. In other words, quantile maps tries to arrange groups so they have the same quantity. As a result, the shading will look equally distributed in quantile types of maps.
- Class 1: 4 – 6 (56 countries have 4, 5 or 6-letter names)
- Class 2: 6 – 7 (38)
- Class 3: 7 – 8 (19)
- Class 4: 9 – 11 (36)
- Class 5: 12 – 24 (27)
Quantile maps takes the total of number of features (176 countries in our case). Then, it divides the total by the number of classes to get the average (176/5=35.2). Finally, quantile maps counts the quantity in each group and arranges them as close to the average as possible.
You can see how the count of each class looks very similar and are close to 35.2. For each class, there are not too many or too few for count.
Despite the balanced style in quantile choropleth maps, they can also be misleading. They are misleading because people tend to look at a shade and group it in the same category. For example, a 12-letter country gets the same dark shading as a 24-letter country… and where’s the justice in that?
Natural Breaks (Jenks) Classification
The first thing to remember about the Natural Breaks (Jenks) classification is that it is an optimization method for choropleth maps. In short, it arranges each groupings so there is less variation in each class or shading.
- Class 1: 4 – 6 (56)
- Class 2: 6 – 8 (57)
- Class 3: 8 – 12 (41)
- Class 4: 12 – 18 (18)
- Class 5: 18 – 24 (4)
Natural Breaks (Jenks) takes an iterative approach by comparing the sum of squared deviations between classes to the array mean. Then, the algorithm uses a goodness of variance fit with 1 as a perfect fit and 0 as a poor fit.
The founder of the Natural Breaks data classification method was a cartographer by the name of George Frederick Jenks. He specialized in monitoring the eye movements of people when looking at a map. And the results for this map looked great too.
You can see how this data classification method minimizes variation in each group. As we have lots of shorter country names, it finds suitable class ranges. But it still manages to group outliers with longer country names in a class of its own.
Standard Deviation Classification
Standard deviation is a statistical technique type of map based on how much the data differs from the mean. You measure the mean and standard deviation for your data. Then, each standard deviation becomes a class in your choropleth maps.
In our case, the mean number of characters is about 8.5 with a standard deviation of 3.7 characters. As a result, all countries with 5 to 8 characters will be placed in the 0 to -1 standard deviation grouping. Likewise, countries with 9 to 12 letters are grouped in 0 to 1 standard deviation range like this:
- Class 1: <-1 σ (9)
- Class 2: -1 to 0 σ (104)
- Class 3: 0 to 1 σ (41)
- Class 4: 1 to 2 σ (10)
- Class 5: 2 to 3 σ (9)
- Class 6: 3 to 4 σ (2)
- Class 7: >=4 σ (1)
The raw categories as output need a bit of clarification to the reader. What is the average? What is the range for each standard deviation?
Despite these inconsistencies, standard deviation types of maps might be one of the most appropriate because of its statistical origin. All the 4 letter countries are <-1 standard deviations. Countries with 5 to 8 letters are -1 to 0 standard deviations. The one 24-letter country is >4 standard deviations because of its extreme deviation from the mean of 8.5.
Pretty Breaks Classification
If you want round numbers in your ranges, then you should choose pretty breaks. All pretty breaks does is rounds each break-point up or down. So instead of having a break point as 599.364 it will become 600,000 with pretty breaks.
It’s a bit hard to see how round the numbers are (it’s grouping by 5’s) in this example because all the examples above also produce round numbers. But when you have large numbers like population estimates (see below), it will generate some very pretty breaks.
- Class 1: 4 – 5 (29)
- Class 2: 5 – 10 (111)
- Class 3: 10 – 15 (24)
- Class 4: 15 – 20 (10)
- Class 5: 20 – 24 (2)
As a result of making rounded numbers, pretty breaks will also be very picky for the number of classes you decide.
Here’s how population estimates compares between the data classification techniques:
Equal Interval:
Quantile:
Natural Breaks (Jenks):
Pretty Breaks. Now that’s pretty:
Try It Out Yourself
Choropleth maps use different shading and coloring to display the quantity or value in defined areas.
Often the case, the map maker uses a type of data classification to produce its own unique choropleth map. Each data classification method impacts the reader differently.
There are several ways to classify data in GIS. We’ve outlined their differences with different examples for choropleth maps. Use this guide to classify practically anything like crime rates, level of education and politics.
What is your favorite data classification method? Let us know with a comment below.