فصلنامه سنجش از دور و GIS ایران
سال دوازدهم شماره 4 (پیاپی 48، زمستان 1399)

Iranian oak has been affected by oak canopy level dieback in recent years. This phenomenon has led to damage a vast part of the oak forests in the Zagros arena. As to the suitable temporal and spacial resolution of the recent satellite images, it seems promising to detect the forest dieback by remote sensing. The spatial capabilities of Spot6 images with pan and spectral resolution of 1.5 and 6 m, respectively in detecting the drying of oak trees was investigated. The forest area was located on ​​Kuhmareh district of Shiraz in Fars province. The values of different indices such NDVI, EVI, TDVI, SAVI, RNDVI, OSAVI, DVI, MSR of each tree stocks was obtained and the corresponding quantity of dryness was determined at the filed. The best correlation was obtained between TDVI and the observed data. A non-linear function was built based on TDVI standard deviation to predict the dryness of more than 30% as y=17.92(x-0.06)-0.32 with an R2 = 82%. Monitoring forest areas to understand the decline or recovery of trees will be of great help to the forest management community. Therefore, using the results of this study can be a proof to compare the current situation with future periods.

Application of the remote sensing methods in crop area mapping on a large spatiotemporal scale serves is as an alternative to costly time-consuming field data gathering methods. So far, some methods have been developed for wheat and rice area mapping using the images from optical and radar sensors. Some of these methods are appropriate for humid climates with several cloudy days, while others use complex processes in terms of combining both optics and radar images. Meanwhile, methods based on the unique variation of the vegetation index time series belongs to each crop are relatively simple methods that can be used for crop area mapping. The objective of this study is to improve one of the proposed methods for rain-fed wheat area mapping, in which a step-by-step elimination algorithm of non-wheat pixels was applied to MODIS images. The Improved algorithm took advantage of both MODIS and Landsat Images in terms of their high temporal and high spatial resolutions, respectively. The mentioned process could detect rain-fed wheat areas from the pastures and heterogeneous areas with higher accuracy in comparison with the previous algorithm. The overall accuracy, Kapa index, and F1 score for the final rain-fed wheat map was 92.5%, 0.67, and 0.71 respectively.

Estimating the biomass values in forests stands through remote sensing is important. It has been reported that the major reasons of uncertainty are the lack of concurrency in satellite data and field information as well as the use of global allometric equations for estimating the weight of biomass of forest trees inside the country. Minimizing the above problems and the investigation of data performance in developing appropriate model for estimating the forest biomass in the Bankoll region of Karazan District of Sirvan County in Ilam province using Sentinel-1 satellite data in 27th of June, 2017 was the main goal of this study. Average size of the trees crown in 53 rectangular plots related to the coppice growth form with dimensions of 30×30 mwhich during 23 may 2017 to 10 June 2017 through applying DGPS by RTK method have been implemented on the ground were entered in the process of estimation the value of biomass. The average harvested field biomass was 10.63 Mg ha-1. After extraction of radar features, those features which had the greatest correlation with the values of biomass were selected using genetic algorithm by two models including K-Nearest Neighbor (K-NN) regression and Support-Vector Regression (SVR), then the most appropriate combination was identified and the biomass values were modelled. Models were validated using 26 test plots. Correlation of features obtained from radar data and the value of biomass indicated that features of VH، Mean VV، Mean VV GLCM (Correlation) and Mean VH GLCM (Dissimilarity) had the greatest sensitivity towards the value of biomass. Using regression models indicated that SVR model (Relative RMSE of 0.08) was more precise compared with K-NN regression (relative RMSE of 0.10). The best combination in the use of K-NN regression model with a relative RMSE of almost 0.99 Mg ha-1 (equal to 10%) and the coefficient of determination (R2) of 0.22 and the best combination when using SVR model was a relative RMSE of 0.87 Mg ha-1 (equal to 8%) and the R2 of 0.14. The results indicated that the final models, obtained from the optimal features extracted from radar data in the wavelength of C band and used parametric and non-parametric regressional methods in this research, were not abled to improve the saturated effect in data for estimation of biomass in the studied forests and it was not resulted in presenting an estimating model with an acceptable accuracy.

The growth of the urban population has been led to increasing of the urban spaces and growth of the city size. as a result of further construction and alteration of the land available to the benefit of its built-up spaces. Special location the city of Urmia at proximity of the Urmia lake and unfavorable condition of this lake reveals the necessity of the proper landuse planning at this city. One of the required tools for proper planning in this field is the use of remote sensing techniques. The present study aims to evaluate these changes (period 1989-2015) and predict its future trend. SVM and neural network methods are used to evaluate changes in 5 classes Due to the high accuracy of the classification of the neural network, the results of this classification have been used to predict changes for the 2045 horizon. Land constructed in 1989 is 7469.1 hectares, reaching 9217.3 and 94366.9 hectares in 2002 and 2015 respectively, and by2045, according to the prediction model, the neural network is equal to 22449.6 hectares, which is built on lands 13012.7 Shows hectares of increase. The determination coefficient (0.73) and rock curve (82.55%) also indicate the high accuracy of the neural network model to predict urban development changes. The Heldern method results shows that all of these constructions are not based on the real needs of the city And the sparse phenomenon has happened.

Land-cover/land-use maps are necessary for monitoring land changes and proper planning for managers in agriculture, natural resources and environment fields each year. The method of field data collection using GPS and land survey is time-consuming and costly. Therefore satellite images which have entire coverage and repetition of collection, low cost and real-time data, are usually used so that land-cover/land-use maps are produced. Accurate mapping using technique suitable for today is a key factor. Although in the past, conventional classification methods have been applied to images such as Landsat, using new satellite images and modern classifiers specially machine learning has been growing recently and their effectiveness in preparing land-cover/land-use maps has been very successful. Another advantage of satellite images is repetitious collection and according to that, vegetation changes through time can be used to differentiate land cover types. The Sentinel-2 satellite with the superiority of a pixel rating of 10 meters is one of the appropriate tools to discriminate land cover types. In the current study, Support Vector Machine and Random Forest classifiers on multi-temporal Sentinel-2 images were used to differentiate land use and crop types of Sanjabi plain in Ravansar and their accuracies were compared. To do so, after sampling, Principal Component Analysis was performed for four dates in crops’ growing season and PC1,2,3 bands of the images were combined. The two techniques were implemented on the layerstacks of PC1,2,3 bands of the images and the training samples. Results of accuracy assessments showed that Support Vector Machine, with overall accuracy of 91.36% and Kappa coefficient of 0.8927, produces a more precise land use and crop map rather than Random Forest method.

Land surface temperature (LST) monitoring has been widely used as one of the most important environmental parameters by the high temporal resolution sensors such as the MODIS sensor (daily temporal resolution capability and spatial resolution of one kilometer). One of the main problems of these sensors is their low spatial resolution, which limits the performance of these sensors for applications such as fire detection in forest areas and the study of urban thermal islands. In contrast, high spatial resolution sensors such as the ASTER sensor (90 meter spatial resolution and 16-day temporal resolution at the land surface temperature product), they have low temporal resolution, which results in application such as rapid change monitoring. In fact, due to technical limitations, there is no sensor that has a high resolution in spatial and temporal dimensions. To solve this problem, low-cost and efficient spatial-temporal fusion methods have been developed. The most important methods for fusion spatial-temporal methods are enhanced spatial and temporal adaptive reflectance fusion model (ESTARFM) and Spatial and Temporal Data Fusion Approach (STDFA). This work uses the ESTARFM and STDFA algorithms and a new method (SWT-STDFA) based on the STDFA method and the two-dimensional stationary wavelet transformation to fuse LST data spatially and temporally. The LST products of ASTER and MODIS sensors were fused for a part of Tehran city and finally, a virtual image was obtained with a spatial resolution equal to that of the ASTER sensor and a temporal resolution equal to that of the MODIS sensor. Also, based on the existence of a classification map prepared on the basis of normalized vegetation difference index (NDVI) in STDFA and SWT-STDFA algorithms, the effect of using normalized Green Difference Vegetation Indices (GNDVI) and soil adjusted vegetation Index (SAVI) on the accuracy of the synthetic image of the output is discussed. The results of the research indicate the high accuracy of the proposed method with the root mean square error of 3.03 Kelvin, standard deviation of 2. 21 Kelvin, mean absolute difference 1.72 Kelvin and correlation coefficient of 0.92 between the image of the actual land surface temperature and the predicted synthetic image Compared to the other two methods. Also, the increase of vegetation’s indices GNDVI and SAVI in the classification of both STDFA and SWT-STDFA methods did not have much effect on the accuracy of the synthetic image of the output.

The presence of dry and green vegetation in pixels containing spectral information is essential in geological and mineralogical studies. Thus, retrieving sub-pixel information, including estimation of a mineral’s quantity in a single hyperspectral RS image pixel is very important. In this study, the vegetation corrected continuum depth (VCCD) method was trained and its results were validated using spectrometry, laboratory mineralogy, and Hyperion image to reduce the effect of vegetation on the estimation of minerals. The study was conducted in Oghlansar region located in northwestern Iran. SAVI and absorption depth (2102 μm) were used for the estimation of the green and dry vegetation, respectively. Meanwhile, the trained models do not have a high sensitivity to the presence of noise in the spectrum and vegetation type changes. The correction of continuum removed band depth (CRBD) analysis was possible up to 60% for maximum green vegetation cover threshold, 56-60% for dry vegetation, and 72-76% for both dry and green vegetation. Effect of noise and different vegetation types on model capability was examined and the result shows that VCCD is not highly sensitive to random noise and changes in vegetation types. After correction of the coefficients and confirmation of its efficiency, the model was used to correct CRBD and reduce the effect of vegetation on Hyperion image. In the estimation of kaolinite and muscovite, the presence of green and dry vegetation led to the underestimation of the minerals present in the study area. The results showed that VCCD was able to increase the prediction accuracy (R2) by 0.25 and 0.13 and reduce RMSE by 0.0108 and 0.125 for kaolinite and muscovite, respectively.