April 18, 2015 The Methods and Models of Drought Monitoring by Meteorological Satellites in China Henan Provincial Meteorological Institute, P. R. China.

The Methods and Models of Drought Monitoring by Meteorological Satellites in China*Henan Provincial Meteorological Institute, P. R. ChinaDr. Chen Huailiang

OUTLINE1.Introduction2. Monitoring methods and Models3. Some improves to thermal inertia 4. Several problems in drought monitoring by meteorological satellites5. The operational support system for drought monitoring6. Ecology & Environment Monitoring and Assessment7. The Development Planning in the Future

1. IntroductionThe drought is one of the major meteorological disasters affecting agricultural production. The monitoring on drought timely is very important to instruct the disaster defence and agricultural management. Real-time monitoring on drought or soil moisture using the remotely sensed data of weather satellite has many advantages such as low costing, macroscopic, objective, real-time etc.

At present, the methods of measuring soil water are still drilling soil, neutron instrument, TDR, FDR etc.. However, these ways have the same disadvantages, that is their monitoring patterns are discrete points, they cant realize the soil moisture monitoring in large scale.Now, there are many kinds of methods and models to monitor the soil moisture using NOAA and EOS/MODIS satellite data, such as the thermal inertia, the crop water shortage index (CWSI), vegetation water supply index (VSWI), the thermal infrared method and so on.

2. Monitoring methods and Models2.1 The thermal inertia The thermal inertia is the present widely-used method in agrometeorological operation and is mainly suitable for the bare land surface. The thermal inertia (P) is defined as following :

In actual application, we often use the apparent thermal inertia (ATI) replacing the real thermal inertia (P) for the purpose of simple calculation. Through constructing the liner equation between W and ATI, we can retrieve the soil moisture (W): W=a*ATI + b Of course, the equation can also be established with the other patterns, such as exponential power and logarithm function etc..

The apparent thermal inertia (ATI) can be calculated by using the following equation Where K stands for the parameter related with soil character; A for full bands reflectivity; Td for soil surface maximum temperature in day, Tn for soil surface minimum temperature in night. Generally, we usually use AVHRR CH4 brightness temperature replacing the Td and Tn.

2.2 Crop water shortage index (CWSI)Where W stands for soil moistureL for evaporation latent energyE is evaportranspirationEP for latent evaportranspirationb1b2 for regression coefficients This method is based on the energy balance equation and mainly uses in the vegetation cover land. It considers many factors and has explicit physics significance. According to the studies, the CWSI and soil water Whave the following relationship:

Vegetation Condition Index (VCI) is defined as the following formula :Where NDVI stands for normalized difference vegetation index; NDVImax for maximum NDVI in the specified period; NDVImin for minimum NDVI in the specified period. The value of VCI is from 0 to 100. The bigger the VCI is, the better the vegetation growth, and the better soil moisture is.2.3 Vegetation Condition Index and Temperature Condition Index

Temperature Condition Index (TCI): Where T stands for surface soil temperature, but T is generally replaced by the brightness temperature of AVHRR CH4. The value of TCI is from 0 to 100. The bigger TCI is, the warmer the soil surface is, and the worse soil moisture is.TCI is defined as the following formula :

The relations VCITCI and drought or soil moisture are:According to the following index table, we can transform VT Index to drought grades:(1) VT is defined by using VCI and TCI according to the following formula:

2The regression models between soil moisture and VCI , TCI, VI areWhere W stands for soil moistureabc for regression coefficientsVCITCIVI for the same as before

2.4 Regression method of single phase data In actual operation work, sometimes its very difficult to gain the clear-sky remote sensing data of multi phases, then the single phase method is put. The regression equation of the single phase method is defined as following:Where a, b, c stands for the coefficient of regression equation; NDVI,CH2,CH4 for the same means as before.

2.5 Vegetation water supply index (VSWI)The smaller VSWI is, the severer the drought is.Where Ts stands for the temperature of the crown layer of crop. VSWI is defined as the following formula:VSWI = NDVI / TS

2.6 Anomaly of NDVI The growing of vegetation is mainly related with the soil moisture under the normal temperature and sunlight condition. The anomaly of NDVI in ten-days may reflect the moisture content supply condition in a way. So ATNDVI is defined to evaluate the drought grade. As following:TNDVI stands for the NDVI value in the current ten-day and the max NDVI value of ten-day period; for the mean NDVI value; t for days.

2.7 Thermal Infrared Method The temperature distribution of soil surface may indirectly reflect the soil moisture. The temperature rises higher in low soil moisture condition than in high moisture. We can calculate the diurnal temperature difference in terms of satellite data, such as CH4 of AVHRR, and get the drought grades. For the complexity of the land surface, the thermal infrared method has low precision generally.

3. Some improvements to thermal inertia method3.1 Establish corresponding soil moisture monitoring model according to different soil types There are different thermal characters to different soil types because they have different reflectivity and thermal inertia etc.. So, we should consider soil types when we use the thermal inertia method. The results show that the precision can increase more than 2% after distinguishing soil types in establishing the retrieve equation between W and ATI.

3.2 Correction model by surface vegetation cover The thermal inertia method cant be used in vegetation cover land because the different temperature cycle variation character. In the case, we can correct the thermal inertia by using NDVI. The commonly correcting function is :The temperature cycle various character in different land cover T stands for the temperature difference after vegetation correction. For convenience, we use T = NDVI100/c, where c for a constant, set to be 10 after examination.

The correlation coefficient of soil moisture content model (linear model or power function) is improved after the NDVI correction, which proved the correction is applicable, and the corrected thermal inertia method can be used in crop growth stages.Soil moisture content remote sensing image of 20cm depth corrected by vegetation cover (Apr. 16th 2003)

3.3 Consideration of wind speed Radiation balance equation is often used to measure soil moisture by the apparent thermal inertia. The key of which is how to calculate the difference of the surface temperature between day and night. Wind influences strongly the heat exchange of land surface and the crop evaportranspiration. Measuring the soil moisture by remote sensing often neglects the wind speed, or considers it as a constant, or requires inputting real-time wind speed. The themes result in the soil moisture can not be measured timely or the precision are lower. So, the wind speed should be considered in the apparent thermal inertia method.

Two parameters (F and R) related with wind speed were defined, they can all be derived form the DEM data. The two parameters show that the effect of the terrain on the wind speed reassigning , and the surface sensible heat flux and latent heat flux indirectly. So they can be introduced in measuring the soil moisture by remote sensing.

The relative error rate of consideration wind speed is lower 0.4%5.3% (the average is 2.0%). The error rate would increase with the soil depth increase. The average falling value only 0.6% in the soil layer of 30cm. This may indicated that the effect of wind speed will weaken when the soil depth increases. The effect of wind speed can be neglected when the soil depth is more than 30cm.

4. Several problems in drought monitoring by meteorological satellites4.1 The difference of drought and soil moisture measuring by remote sensing Soil moisture is only referred to the water amount in the soil. But the drought is often referred to the shortage of soil moisture during the growth stages of crop. So the Drought is a agrometeorological concept and its monitoring should be aimed at the special crop and special growth stage. The soil types should be considered also at the same time. We got the drought index of Huang-Huai plain of China through field experimentation.

Drought Index in different crop and growth stages in Huang-Huai Plain of China

Drought Monitoring Map in Huang-huai Plain in China by NOAA/AVHRR Data

4.2 The best depth of drought monitoring by weather satellites The best depth of drought monitoring by weather satellites is about 20cm. There is no accordant best depth at present, but it is sure that the depth is not too deep. The reason is that the daily range of soil temperature between day and night is various in different soil layer. The daily range of soil surface temperature is the largest and the deep layer is smaller correspondingly. The daily range of temperature will become zero when the deep comes to some depthit is called : the disappearing layer of daily range.

4.3 The deep soil moisture monitoring by remote sensing The soil moisture changes from surface to deep layer is non-linear tendency, the model that soil moisture at depth according to the surface soil moisture is proposed:

By the support of GIS, the soil moisture of 10cm can be firstly retrieved with the thermal inertia method. Then the soil moisture value of other layers can be calculated by above model. By this kind of method , the soil moisture of any soil layer in 100cm depth can be calculated , and the error can be controlled in 20% or so.

4.4 the model of relationship between soil moisture and the thermal inertia The accuracy of drought monitoring by remote sensing is affected by the models suitability between soil moisture and the thermal inertia. The linear model is one of the models that are wildly used:Other models: Exponent model Function model Logarithm model W stands for the soil moisture, a and b for coefficient , P for the thermal inertia, ATI is the apparent thermal inertia.

However, it still is a difficult and hot problem that monitoring drought by meteorological satellite. Some new techniques (such as microwave sensing), new methods (such as 3S integrated techniques), new models (such as some dynamical models) are developing. With the accumulation of data and the deeper of research, the difficulty of drought monitoring can be solved to a certainty.

5. The operational support system for drought monitoring5.1 The hardware and software The meteorological satellite receiving system for NOAA, FY-1, and the DVB-S of EOS/MODIS ; GIS and many remote sensing softwares; In addition, there are GPS , microcomputer work station , Pentium computers and digitized apparatus, scanner , plotter , colored laser and jet ink printer ,etc.

Hardware Configuration ChartData Flow Chart

The receiving system of NOAA/AVHRRFY-1

The DVB-S receiving system for EOS / MODIS Data

The software used in researchErdas8.7ENVI4.0Developed by our staffs

The GPS apparatus from GARMIN company of U.S.AGPS12XLCDGPS2000-ST2020S

GPS data management system-Mapsource 4.0

The GIS softwares used in workARCVIEWCITYSTARMAPINFOARCINFO

Drought monitoring based on " 3S " technology

5.2 The operational service on drought monitoringWe issued termly drought monitoring information through the report of Henan Provincial Agrometeorological & Remote Sensing Information.

The Remote Sensing Monitoring Map on DroughtWe began the research of remote sensing monitoring on agricultural drought from 1996. At the present, the operational system of remote sensing monitoring on crop drought is becoming perfect. The system can provide remote sensing monitoring service yearly, and the accuracy is over 80%.

Serving for the Economic Construction and Peoples Lives by Means of Newspaper, TV, and Internet

6. Ecology&Environment Monitoring and Assessment

Issuing Report of Ecology & Environment Monitoring and Assessment Service We issued regular or irregular The Monitoring and Assessment Report on Henan Provincial Ecology & Environment from 1 Jan. 2004 based on the remote sensing technology. The contents include mainly drought, flood, sand storm, snow, heavy fog, vegetation, reservior change, land use/ land cover etc.

On the uniform requires of China Meteorological Administration (CMA), from autumn of 2005, we began to issue Henan Provincial Ecology & Environment Climatic Assessment, which quarterly or yearly gave a grade assessment based on climatic humid index, vegetable cover index, desertifized land index, water index and natural disaster index. Another, we also issued the following monitoring services by using remote sensing data and GIS:

LUCC monitoring

Desertified land monitoring

Withdraw infield and return forest monitoring

Wheat growth monitoring

Flood monitoring

Heavy fog monitoring

Forest distribution monitoring

Reservior area change monitoring

Sand storm monitoring

Snow area monitoring

Highway Environment assessment

Forest fire monitoring: NOAA 14: 2001 312 16:45 () K 1 1 2.43 114.29 36.12 301.0 2 1 3.21 113.04 34.83 301.3 3 1 4.13 111.04 34.46 302.8 4 1 2.37 112.65 34.31 301.0 5 1 2.88 112.42 33.28 302.9 6 1 3.29 112.28 33.24 302.3 7 1 1.00 111.08 33.44 301.3 8 1 1.00 111.70 33.31 302.8 9 1 1.00 112.27 33.24 302.0 10 1 1.00 112.41 33.28 302.9 11 1 1.00 113.50 32.53 303.5 12 1 1.00 113.50 32.53 305.4 13 1 1.00 114.14 36.13 305.1

Straw burning monitoringFiringSmogFired area

The provincial net of ultraviolet radiation, soil moisture, atmosphere dry sedimentation etc. have established. The atmospheric component station of Zhengzhou is constructing. They can provide the base data for ecology & environment monitoring and assessment. CO2 MonitoringUltraviolet Monitoring

7The Development Planning in the Future Well set up the information service system of Henan provincial ecological environment and resource monitoring of in the future of 3 to 5 years, which is based on the integrated application technology of 3S. It mainly includes the contents of agriculturedisasterecology & environment and resource ect. four aspects, 20 items. Based on the dynamic remote sensing monitoring, we will develop the assessment and forecasting service step by step by using of the polar orbit meteorological satellite data, EOS/MODIS data and land satellite data. Meanwhile, the perfect annual service scheme will be set up.

X. 25Internet 4 X. 25Internet

Flood

Drought

Crop GrowthVegetation and ForestDisasterAgricultyreEcology&EnvironmentResourcethe information service system of Henan provincial ecological environment and resource monitoring Seeding AreaForest FireCrop ill& pestHeavy fogMountain coastSnow

Frost injure

Earthquake portentWater pollutionSand stormStraw setting on fireDesertified LandCity hot islandLand use / land coverLand of urban constructionWater resourceThe system structure chart

The Elementary Information Service System of Ecological Environment and Resource Monitoring

3D Topographical Map of Henan province