H26+V5+Shaanxi

This area is in Shaanxi and appears to have some positive growth and I want to look into whether this growth is related to a decadal rainy season that alternates around China.

=Landsat Data=





I decided to do EVI because I was afraid of saturating the image using NDVI. There has generally been a lot of positive growth with in this region and the fear of saturating the data especially during summer was a valid concern. During the summer, values are very high representing the growing season and shows very saturated vegetation regions. The northwest region shows very light vegetation while the east and southern regions show the heavier vegetation.
 * The images that I used were days 153 (June 2nd), 233 (August 21st), and 289 (October 16th).

** Lab 4 **
 * Question 1.** Write down the image dimensions (e.g. number of lines, columns and bands) for the Anthromes file and the MODIS time series.
 * File type || Rows || Columns || Bands ||
 * Anthromes File || 121 || 303 || 1 ||
 * MODIS Time Series || 201 || 504 || 144 ||


 * Question 2.** How do the number of lines and columns compare for the time series and the Anthromes file?Are they the same or different, why? How many time steps does your time series have? How many observations per year?

The two files have different amounts of rows, columns and bands. The MODIS time series has more rows columns and bands corresponding with the bigger picture image. The time series has 9 steps and 16 per year.


 * Question 3.** Discuss why it would not be right to shrink the Anthrome pixels.

It would be unwise to shrink the Anthrome pixels because the Anthrome pixels are classification data while the MODIS pixels are float pixels. The MODIS pixels would be easier converted to the Anthrome pixel size because shrinking an image creates difficult half bytes.
 * **#** || **Class** || **#Pixels in Area** || **%Pixels in Area** ||
 * 11 || Urban || 236 || .646 ||
 * 12 || Dense Settlements || 1357 || 3.714 ||
 * 21 || Rice Villages || 401 || 1.097 ||
 * 22 || Irrigated Villages || 3210 || 8.784 ||
 * 23 || Cropped Pastoral Villages || 996 || 2.726 ||
 * 24 || Pastoral villages || 1649 || 4.513 ||
 * 25 || Rainfed Villages || 2129 || 5.826 ||
 * 26 || Rainfed Mosaic Villages || 875 || 2.395 ||
 * 31 || Residential Irrigated Cropland || 654 || 1.79 ||
 * 32 || Residential rainfed Mosaic || 3823 || 10.462 ||
 * 33 || Populated irrigated cropland || 38 || .104 ||
 * 34 || Populated rainfed cropland || 435 || 1.190 ||
 * 35 || Remote Croplands || 8 || .022 ||
 * 41 || Residential rangelands || 5041 || 13.795 ||
 * 42 || Populated Rangelands || 7255 || 19.854 ||
 * 43 || Remote Rangelands || 6056 || 16.573 ||
 * 51 || Populated forests || 741 || 2.028 ||
 * 52 || Remote Forests || 105 || .287 ||
 * 61 || Wild Forests || 0 || 0 ||
 * 62 || Sparse Trees || 0 || 0 ||
 * 63 || Barren || 1466 || 4.012 ||


 * 5 Most Dominant Classes**


 * 1) Populated Rangelands
 * 2) Remote Rangelands
 * 3) Residential Rangelands
 * 4) Residential Rainfed Mosaic
 * 5) Irrigated Villages



NDVI (Left Image) Coefficient of Variance (Right image)

69.468%
 * Question 4.** How much area is covered in your region by the 5 most dominant classes.


 * Question 5.** Discuss the differences between NDVI and the coefficient of variation for the 5 classes.

The Residential Rainfed Mosaic and the Irrigated Villages have the highest NDVI values which is to be expected as the land is fed by rain and irrigation keeping plants healthy and very reflective in the NIR region in this area. The other three classes have much more defined peaks and much clearer points for the start and the end of the growing season than the Residential Rainfed Mosaic and the Irrigated Villages. This corresponds to the high level of the Coefficient of Variance shown in the graphs for the Residential Rainfed Mosaic and Irrigated Villages.


 * Question 6.** Can you determine when the growing season starts and ends for each class? Does it change by year?

In general it's fairly easy to spot the growing season. The Remote Rangelands, Populated Rangelands, Residential Rangelands and Residential Rainfed Croplands all have a fairly distinct start and end to the growing season. Because of the satellite data, there is no available data for the last past of the 2008 year. The irrigated villages NDVI has a very different yearly progression. It has a slow progression as if it were the start of the growing season, but then there is a slight dip in the NDVI values before picking up and peaking in the late summer period. The growing season appears to happen around the same band of the each year (band 4) or sometime in March.


 * Question 7.** How do you think the phenology in this biome would have looked if there were no humans?

If there were no humans in this region, the biome would be greatly dominated by the Remote Rangelands classification. The NDVI values for the region would be generally a lot lower which would correspond to the lower amount of vegetation in the region that is associated with rangelands.


 * Question 8.** Do you find anything remarkable in the time series? Are there trends?

In general for all 5 NDVI time series, there is a slight increase in the NDVI peaks and an increase from the previous year of the minimum NDVI values. As a whole, the NDVI values appear to be fairly stagnant with no overall trends.


 * Question 9.** Are there other things you want to discuss in these graphs?

The Irrigated Villages NDVI time series is the most interesting and the sudden drop in NDVI values throughout the time series is something to investigate. This could be something to due with crops that are being grown.

=

Lab 5=

For lab 5, I wanted to look at MODIS temperature data to determine if there was a correlation between the small vegetation growth trend and the temperature. I started out downloading the Temperature data from MODIS for the period of July which is about the maximum of the growing period according to the NDVI graphs from Lab 4. Using layer stacking, I created a time series for the July temperature from 2000-2009. I then used band math to change the temperature from the already corrected Kelvin to Celsius by subtracting 273.15 from the values given from the MODIS data. I created a density slice for the day time temperature with minimum values of -10 degrees Celsius and maximum values of 55 degrees Celsius. The high extremes for temperature are necessary because the time period is in the middle of the summer and much of the north western corner of the area is covered by desert. The desert can cause higher temperature extremes for MODIS data. The temperature seemed to decrease in general from the start of the decade towards the end of the decade. To qualify this, the average of the time period for each year was taken using ENVI and then plotted via excel.
 * [[image:DR_5_2000_Temp.jpg]] || [[image:DR_5_2009_Temp.jpg]] || [[image:DR_5_Legend.png]] ||
 * = 2000 ||= 2009 ||  ||



This trend verified the lower temperatures that were being witnessed directly from the Time Series. The increased vegetation increase could be in direct cause of the cooling temperatures. Cooler temperatures during the summer could cause decreased killing of vegetation would keep NDVI values higher. Also, there is a decadal oscillation around China that could cause this to happen. Cooler decadal temperatures can cause wetter climates which would also help the vegetation growth. It's possible that this region of China is experiencing this decadal shift.