CC BY
10
НАУКИ О ЗЕМЛЕ
PASTURE MONITORING BY MODIS SATELLITE DATA_
Khandsuren D2
Mongolian State Agricultural University, School of ecological and technology development
Gerlee.Sh2
Department of land management, School of Agroecology
Tamizhazhagan.V3 Mongolian University of Life Sciences Department of Zoology Annamalai university in India
ABSTRACT
Monitoring of pasture vegetation in all territory of Mongolia was made on the basis of NDVI (Normalized Difference Vegetation Index) data by remote sensing method or MODIS satellite. NDVI (Normalized Difference Vegetation Index) data, which are product MOD 13 of MODIS satellite, are used for rangeland vegetation and drought index of our country. Although vegetation growth period is shorter in Mongolia, mapping of pasture vegetation pattern and analysis of general conditions of summer at 16 days intervals from May through September of each year is important for optimal management of movements for good pasture in our country's animal husbandry depending on pasture vegetation or drought situations. MODIS/NDVI (Normalized Difference Vegetation Index) data were analyzed within the framework of "National Geo -Information Center for natural resource management" project, a total of 260 data were used for evaluation of pasture vegetation and the present study was performed according to methods of monitoring.With the purpose of comparing multi-year average NDVI values in May to August in 2005-2013 with those NDVI values in 2013, statistical calculations were made and data were demonstrated with relevant aimags and soums. As well, increase and decrease of NDVI values in 2013 as compared to multi-year values was investigated. Methods of evaluation and regular monitoring of pasture vegetation biomass were described and activities intended for confirmation and improvement of the methods by land measurement data were implemented.
Key words: vegetation index, satellite data, condition index, pasture
INTRODUCTION
As a result of installation of MODIS data station, it formed opportunity to assess and map, two times daily, pasture vegetation condition of Mongolian wide territory, natural disasters such as drought, snow cover, dust storm, forest fire and every year natural resource such as forest and surface water [6].
MODIS satellite data has capacity to identify space of 250 meters to 1 kilometer by 36 canals of spectrum; the capacity is 16 times more than NOAA satellite. This satellite data is very significant to map natural disaster such as drought, dzud, forest fire and steppe fire, natural resource, land cover, water resource of forest and surface and pay control on its change.
The purpose of the research is to monitor pasture properties of Mongolian territory on the basis of NDVI (Normalized Difference Vegetation Index) data by remote sensing method MODIS satellite, evaluate change in comparison with multi-year average , analyze in combination with other data, calculate drought index and process methodology to make pasture monitoring. For these goals, the following objectives will be used:
1. To make and finalize mapping method to monitor pasture properties on the basis of data of MODIS satellite.
2. To map Mongolian territory and pasture properties.
3. To calculate vegetation biomass by vegetation drought index.
Satellite data to be used in the research
The research used NDVI data of MODIS/TERRA group of satellite per 16 days for 8 years from 2005 to 2013 which processed and globally integrated by National Aeronautics and Space Administration (NASA) of USA.
MODIS satellite data has constant frequency, easy to use, useful to permanently map, assess and monitor earth surface by wide range of 2300 km and 36 canals of light spectrum and use for the research.
METHODOLOGY
In the frame of research, as a result of using remote sensing method and technology for pasture vegetation monitoring, it will make pasture and vegetation map by MODIS satellite with purpose of daily meteorological and natural data and validate satellite data as a result of ground pasture biomass data.
Normalized Difference vegetation Index (NDVI)
Because NDVI data is data which converted to integer multiplied by 10000, it is necessary to convert -1 to+1 fraction number. This index is calculated comparing difference of light reflection value which measured in NIR and RED with its addition.
For example:
NDVI = (NIR-RED) / (NIR+RED) (1)
Here:
NDVI- Normalized Difference Vegetation Index/
NIR- near infra red
RED- infra red
NDVI data of green flora is fluctuated between 0 and 1; non -green flora such as cloud, snow, rock is fluctuated by fraction number between 0 and -1. Higher plus value, higher flora vegetation.
Covering territory of Mongolia: from southern latitude 411 to northern latitude 531 and western longitude 80 5to eastern longitude 120 1
The covering data:1500 rows and 2500 columns NDVI data is integer number with precision of 8 bits (0-255), NDVI data is converted to fraction number with the following formula.
NDVI = $n1_mod13q1/10000.0 (2)
From the above-mentioned data, data relating to Mongolian territory was cut and used for the research. Research processed satellite data by ERDAS IMAGINE 9.2 software of image process and ARCGIS 9.2 software of ground data system. In the research, it used methods such as multi-dimensional mathematical statistic method, time series analysis, and comparison and change evaluation of naked eye observation.
RESULTS
Whole territory of Mongolia
On the basis of daily and multi-year archive data of NDVI. It is showing figure of time and space change of pasture vegetation condition.
Figurel. NDVI - August 28 (2005-2013)
As August 28, between and 2005 and 2013, NDVI Figure 1 has not any additional processing, it attempted to determine change of vegetation value and summer vegetation by naked eye analysis. In seeing from figures, vegetation value of central and eastern region of Mongolian territory as 2013 is higher than other years. Multi-year average value of NDVI By ERDAS 9.2 software, it is possible to cut data of province and soum, make all processing of the first
phase and it researched change of MODIS/NDVI data for 8 eight years by all province and territory of Mongolia. To process data, it divided provinces by province territory, calculated multi-year meaning by 16 days, and compared the data. Table 1 made statistical calculation to compare NDVI value as 2013 with multi -year NDVI value as May to August between 2005 and 2013, result was shown by each province [5].
By territory of all province of Mongolia
Table 1 Multi- year NDVI value and as May to August between 2005 and 2013
1 Arkhan gai Bulgan Khuvsg ul Uvurkha ngai Selenge Tuv Khovd Uvs Bayan-ulgii Govi-altai
145 0.328 0.418 0.374 0.173 0.489 0.318 0.136 0.176 0.179 0.101
161 0.41 0.51 0.473 0.193 0.596 0.371 0.159 0.208 0.229 0.114
177 0.488 0.502 0.552 0.228 0.628 0.424 0.175 0.226 0.282 0.127
193 0.532 0.595 0.583 0.252 0.661 0.445 0.18 0.235 0.29 0.132
209 0.532 0.606 0.589 0.276 0.658 0.447 0.181 0.24 0.281 0.133
225 0.506 0.595 0.555 0.271 0.639 0.447 0.177 0.241 0.264 0.133
241 0.446 0.533 0.485 0.25 0.596 0.258 0.167 0.226 0.224 0.13
i Bayankho Zavkha Dorno Khenti Sukhbaat Dundgo Umnugo Dornogo Averag
ngor n d i ar vi vi vi e
145 0.12 0.235 0.268 0.331 0.206 0.119 0.083 0.144 0.233
131 0.138 0.288 0.318 0.38 0.231 0.124 0.089 0.115 0.275
177 0.164 0.32 0.348 0.434 0.284 0.144 0.094 0.127 0.311
193 0.177 0.337 0.401 0.478 0.324 0.151 0.097 0.134 0.334
209 0.186 0.345 0.436 0.481 0.332 0.163 0.104 0.143 0.341
225 0.182 0.334 0.437 0.482 0.339 0.18 0.111 0.157 0.336
241 0.166 0.295 0.413 0.439 0.316 0.174 0.115 0.161 0.308
It calculated Normalized Difference Vegetation Index of MODIS per a 16 days for the period from May 24, 2013 to August 28, 2013 by all provinces of Mongolia, average meaning was 0,233-0,30, value of Gobi-
desert region, Gobi desert and steppe region is lower than average meaning.
In comparison with multi -year NDVI value with NDVI value as 2013, certain increase and decrease is determined. NDVI for the period from May 24 to June 09, 2013 was lower than multi-year average by 0,0160,018.
Between June 25 and July 27, NDVI value was increased by 0,003-0, 04 in comparison with multi-year average. Vegetation index as 2013 was intensively increased for the intensive period of vegetation growth from end of June to end of July. This growth may be depended on weather precipitation.
In comparison with multi-year average with NDVI value as 2013, NDVI value of provinces was different with regional difference. For all provinces, summer vegetation as 2013 was better.
But it was observed that summer vegetation of Gobi region, western and eastern region was not better in June and July. Especially, western provinces had less precipitation and plant growing condition was not
appropriate by summer vegetation period of 2013 till July 20.
To calculate vegetation condition index
Due to climate and weather change, natural flora and plants interrupt and dry during growth process phases, normal growth of plant changes. Vegetation condition index is calculated to evaluate vegetation condition.
VCI(ij ,t)={ [NDVI(i,j ,present)-NDVI-min(ij ,t)]/[NDVImax(i,j ,t)NDVImin(i,j ,t)]}*100 (3) Here; VCI (i,j,t) - vegetation condition index This index determines vegetation condition by each 16 days.
If: VCI= 1-30, bad (brown yellow) VCI= 40-60, normal (yellow) VCI= 70-100, good (green) 70-100, good (green) For example:
It calculated comparing pasture condition as August 28 for 16 days since June 09, 2013 multi-year average, max and min NDVI value (Figure 3). Summer vegetation got better from the 2nd ten days of 2013 by western and Gobi region and from the 3rd ten days by central region, from the 1st ten days of July by central region and the 3rd ten days by eastern region summer vegetation got better [4].
Remote sensing drought index
Remote sensing drought index is calculated by satellite data. Drought index is abbreviated with RSDI (remote sensing drought index). RSDI is calculated with following formula.
RSDI(i,j ,t)={ [NDVI(max)-NDVIogoo]/[NDVImax-NDVImin]}*100 (3)
If: RSDI< 20, summer vegetation (dark green) RSDI = 60-79, less drought (brown)
RSDI= 20-39, good (pale green) RSDI > 80, drought (yellow)
RSDI= 40-59, normal (pink)
It was observed that if asJune 09, 2013, incompar-isonwithotherprovinces, westernprovinceshad-
moredroughts, central region had good vegetation. As June 25, some soums of Umnugovi province as Gobi region and eastern provinces had intensive drought. But between 11 and 27 July, drought was being continued due to constant precipitation [5].
By data 12 of August, drought intensity increased by provinces other than central provinces. By end of August, drought intensity decreased by eastern and Gobi region.
CONCLUSION
1. The research was performed by the methodology as a result of processing methodology to evaluate pasture vegetation properties and make monitoring method on the basis of the collected data during the research.
2. It attempted to make statistical calculation to compare average meaning of NDVI which collected for the period from May to August between 2005 and 2013 with NDVI average meaning of 2013, make comparison on each province and soum. In comparing multi -year NDVI progress with NDVI meaning of 2013, increase and decrease progress was studied.
3. On the basis of drought index, it determined that NDVI meaning is low and constant in territory of Umnugovi province as representation of Gobi region and NDVI is higher and lowly - changed in territory of
Arkhangai province. NDVI progress showed that the most part of territory of Dundgovi province is consisted of desert steppe.
Reference
1. Amarsaikhan. D "Using Geographic data system, remote sensing survey for natural resource management" 2006, 23-28 p.
2. Adiyasuren. Ts, Erdenetuya. M "Monitoring vegetation cover of Mongolian benchland" 2003, 23-29p.
3. Use of remote sensing survey and geographic data system, "Compilation of Scientific publications of National second Conference" 2007, 80- 87 p.
4. Khandsuren. D, " Using Agricultural Meteorological Data in Environment", "Use of remote sensing survey and geographic data system"Compilation of the 2nd Conference", UB, 2007, 24-26page.
5. Khandsuren. D "Using MODIS, FY-2C satellite data in pasture monitoring", 2009, Nederland's ITC, 2008, 65-75 page.
6. Erdenetuya. M "Remote sensing methodology and technology of pasture monitoring"Dissertation, UB, 2004
7. www.edsimwww.cr.usds.gov/pub/imswelco
me
