h22+v11

This section contains the southern half of Madagascar. Over the past few decades, Madagascar's precipitation has decreased as much as 10 percent, and its temperature has risen 10 percent. This is causing long periods of extreme drought and suffering for the local people. Many of the thick tall forests that once covered the land are now decreasing rapidly, and some are gone. I want to see (1) which spots, if any, have decreased so much to be considered desert; (2) find where the wettest or greenest places are; (3) locate urban areas; (4) locate farmland.


 * Here are some interesting facts about Madagascar: **

> -- The World Wildlife Fund -- The World Food Program > -- The World Food Program > -- UNICEF //Source: Los Angeles Times, 2009, []//
 * **Madagascar has lost 90% of its forests.**
 * **In three of the last five years, southern Madagascar has seen crop failures because of drought.**
 * **In the driest regions in southern Madagascar, 1.5 million people have inadequate food because of prolonged drought -- 250,000 of them children younger than 5.**
 * **More than 8,600 children were treated for malnutrition in southern Madagascar from May to October, 2009 because of drought and crop failure, more than double the number expected.**



=MODIS DATA=

These images are from spring, summer and fall in southern Madagascar. I used NDII, NDVI, and NDWI because the region has lost about 90% of the rainforest vegetation and is in drought

conditions. The central area of southern Madagascar seems to be effected the most by the climate changes.

//**Comments from Kirsten: interesting approach to use different indices for different months. I am not really sure how to compare these different values for the different months. Typically we try to use the same index for all months so that we can investigate the time series.**// Spring 2010 NDII

 Summer 2010 NDVI

 Fall 2010 NDWI

LAB 4
Anthromes file – Columns: **184** Lines: **121** Bands: **1** MODIS TimeSeries file – Columns: **305** Lines: **202** Bands: **120**

The TimeSeries file has more columns and lines because it is at a finer resolution than the Anthromes file. The time series has 16-day steps, and 22 observations per year.

You cannot shrink the coarser resolution to the finer resolution because it will cause distortions in the image, and it could also contain incorrect values due to distortions. It is much better to enlarge the finer resolution to the coarser resolution so that the values can be kept correct and so it doesn’t contain distortions.

** Study Class Statistics: **

 * = //#// ||< //Class// ||= //# pixels in area// ||= //% pixels in area// ||
 * = 11 ||< Urban ||= 5 ||= 0.023 ||
 * = 12 ||< Dense settlements ||= 19 ||= 0.086 ||
 * = 21 ||< Rice villages ||= 0 ||= 0 ||
 * = 22 ||< Irrigated villages ||= 0 ||= 0 ||
 * = 23 ||< Cropped pastoral villages ||= 0 ||= 0 ||
 * = 24 ||< Pastoral villages ||= 22 ||= 0.099 ||
 * = 25 ||< Rainfed villages ||= 2 ||= 0.009 ||
 * = 26 ||< Rainfed mosaic villages ||= 99 ||= 0.447 ||
 * = 31 ||< Residential irrigated cropland ||= 43 ||= 0.194 ||
 * = 32 || **// Residential rainfed mosaic //** ||= 909 ||= 4.105 ||
 * = 33 ||< Populated irrigated cropland ||= 6 ||= 0.027 ||
 * = 34 || **// Populated rainfed cropland //** ||= 345 ||= 1.558 ||
 * = 35 ||< Remote croplands ||= 14 ||= 0.063 ||
 * = 41 || //** Residential rangelands **// ||= 836 ||= 3.775 ||
 * = 42 || **// Populated rangelands //** ||= 617 ||= 2.786 ||
 * = 43 ||< Remote rangelands ||= 88 ||= 0.397 ||
 * = 51 || //** Populated forests **// ||= 312 ||= 1.409 ||
 * = 52 ||< Remote forests ||= 47 ||= 0.212 ||
 * = 61 ||< Wild forests ||= 0 ||= 0 ||
 * = 62 ||< Sparse trees ||= 0 ||= 0 ||
 * = 63 ||< Barren ||= 0 ||= 0 ||

**Top 5 Dominant Classes:**
1)  Residential rainfed mosaic -- 909 --  4.105% 2)  Residential rangelands -- 836 -- <span style="color: #0000ff; font-family: 'Arial Black',Gadget,sans-serif;">3.775% <span style="color: #000000; font-family: 'Arial Black',Gadget,sans-serif;">3) <span style="color: #2fac1e; font-family: 'Arial Black',Gadget,sans-serif;"> Populated rangelands <span style="color: #000000; font-family: 'Arial Black',Gadget,sans-serif;">-- <span style="color: #ac201e; font-family: 'Arial Black',Gadget,sans-serif;">617 <span style="color: #000000; font-family: 'Arial Black',Gadget,sans-serif;">-- <span style="color: #0000ff; font-family: 'Arial Black',Gadget,sans-serif;">2.786% <span style="color: #000000; font-family: 'Arial Black',Gadget,sans-serif;">4) <span style="color: #2fac1e; font-family: 'Arial Black',Gadget,sans-serif;"> Populated rainfed cropland <span style="color: #000000; font-family: 'Arial Black',Gadget,sans-serif;">-- <span style="color: #ac221e; font-family: 'Arial Black',Gadget,sans-serif;">345 <span style="color: #000000; font-family: 'Arial Black',Gadget,sans-serif;">-- <span style="color: #0000ff; font-family: 'Arial Black',Gadget,sans-serif;">1.558% <span style="color: #000000; font-family: 'Arial Black',Gadget,sans-serif;">5) <span style="color: #2fac1e; font-family: 'Arial Black',Gadget,sans-serif;"> Populated forests <span style="color: #000000; font-family: 'Arial Black',Gadget,sans-serif;">-- <span style="color: #ac201e; font-family: 'Arial Black',Gadget,sans-serif;"> 312 <span style="color: #000000; font-family: 'Arial Black',Gadget,sans-serif;">-- <span style="color: #0000ff; font-family: 'Arial Black',Gadget,sans-serif;">1.409%

Top 5 classes cover an area of: **// 234,444,878,962.1283 m2 // //OR// // 234,444.878 km2 //**
==== Maximums in the NVDI correlate with the minimums in the CoV, and vice versa. There is a spike in the values for CoV in each class, at around the second year in the sequence. The NDVI’s show no spike in values like the CoV’s do. ====

==== If no humans were in this area, the phenology would likely look similar because most areas of southern Madagascar are sparsely populated. Except for 4 or 5 urban areas, most of southern Madagascar has low population densities. The changes in this region are mostly caused by natural environmental changes, like drought. ====

==== There is nothing remarkable shown in the time series. The only trends I see are that the wet and dry seasons are clearly defined within each year. Also, the time series shows that the wet season generally decreased from 2003-2009. ====


 * //[Figure 1: Stats for Class: Residential rainfed mosaic]//**

Mean

Coefficient of Variation


 * //[Figure 2: Stats for Class: Populated rainfed cropland]//**

Mean

Coefficient of Variation


 * //[Figure 3: Stats for Class: Residential rangelands]//**

Mean

Coefficient of Variation


 * //[Figure 4: Stats for Class: Populated rangelands]//**

Mean

Coefficient of Variation


 * //[Figure 5: Stats for Class: Populated forests]//**

Mean

Coefficient of Variation

LAB 5 : Evaluating NDWI Values
<span style="color: black; direction: ltr; display: block; font-family: Calibri; font-size: 14pt; margin-bottom: 0pt; margin-left: 0in; margin-top: 0pt; text-align: left; unicode-bidi: embed;">The images analyzed are from mid-February, 2000-2011. Over the past 12 years, Madagascar has received much less rainfall than average. The NDWI data shows that most areas of the island have received the least rainfall in years 2003-2005. The last year, however, has received very little rainfall so far. It looks as if the lasting droughts on the island will continue, so current and future vegetations are at risk.

This is centered on a rainforest in the region. NDWI values here are much higher than on the west and east coasts and nearly everywhere else on the island. Spots containing positive NDWI values like this are rare, indicating that most of the island has received very little rainfall in the years examined. The amount of rainforest in southern Madagascar is much less than expected.

The west coast of Madagascar has been going through periods of extreme drought or very little rainfall. This is evidence that this area has received much less rainfall than needed. This Z-Profile shows that this area has extremely low NDWI values, especially compared to the spot of rainforest just to the east.

The east coast of the island has also been receiving low rainfall amounts, although more than the west coast. This shows that this area also has low NDWI values, but not as low as the west coast. Drought conditions here are not as severe as to the west.

=LANDSAT DATA=