Bangladesh is a riverine country with several estuarial islands, these islands are subjected to dynamic erosion and accretion. Increased sea level could consume about 26% of land in low lying zone like Bangladesh by 2100. In this alarming situation, monitoring coastline dynamics is obligatory to assess the current status of coastline so that essential measures can be taken to compensate the damage. In this study remote sensing and geographical information system (GIS) techniques have been used to quantify the erosion and accretion of land along the coastline of Moheshkhali Island, Bangladesh. Satellite images form Thematic Mapper (TM) has been used from 1996 to 2016. Modified Normalized Water Index (MNDWI) algorithm has been applied to separate land from water and on screen digitization has been adopted to extract coastline form images of 1996, 2006 and 2016 respectively. Finally, overlaying the digitized maps, the erosion and accretion has been quantified. The result shows that the western and the south-eastern part of the island faced total accretion of 4830 hectares. Erosion on the other hand was not so dominant in the study area during the 20 year time span. The finding contradicts previous works done for estuarine islands as estuarine island face both erosion and accretion. The root of this anomaly is engraved in a tidal swamp forest named “mangrove”.
Knowledge about land use and land cover (LU/LC) is necessary to plan, monitor and evaluate the developmental activities. Many methods of remote sensing technologies have been developed to detect LU/LC. Several digital classification approaches were used to classify the satellite dataset. Different methods have different principal algorithms and have different approaches. These technologies are required to be evaluated for better accuracy. In the current study, five techniques (isocluster unsupervised classification, maximum likelihood supervised classification, principal component based classification, spectral angle mapper based classification and decision tree approach based classification) have been used to identify LU/LC in Bhubaneswar city in India. The accuracy assessment is done on all the classification techniques to compare and choose the best suited algorithm. It was observed from the study that Decision Tree Classifier based classification showed best result with overall accuracy of 88.40 % and kappa value 0.855 thus suited well to the study area.
Nonpoint source (NPS) pollution models have shown promise in evaluating effects of conservation practices (CP) before these practices can be implemented in the field. An increasing trend has been noted in use of NPS pollution models for assessing effectiveness of CPs. The increase in use of NPS pollution models necessitates their continuous development and improvement. Model improvements will help assess water quality impacts of CPs more accurately. Apart from assessing effectiveness of different CPs, models can also help in identifying suitable location for targeting CPs in a watershed. However, there are challenges associated with modeling CPs such as selecting the best type of CP, best location of CP placement, reliable calibration of the model, validation of modeling results with observed field data, and proper documentation of the modeling process. This paper discusses hydrologic/water quality models used to simulate various CPs in a watershed along with the associated challenges and potential future directions for NPS modeling.
Quick and accurate quantification of lake water quality (WQ) is essential for its management and improvement. Use of geotechnology (remote sensing, GIS, and GPS) applications is a step forward in improving our ability to effectively quantify and manage the WQ of ungauged lakes. Beaver Reservoir, a drinking water source for over 280,000 people in northwest Arkansas, is facing increased chlorophyll-a (Chl-a) and suspended matter (SM) content in the lake. This study is designed to qualitatively predict the Chl-a and SM content in the lake on a spatial basis from Landsat-TM image digital information. A Learning Vector Quantization (LVQ) classification neural model was used to predict the qualitative (oligotrophic, moderately oligotrophic, and mildly trophic) classes for several spatial positions in the lake. The geostatistical tool in ArcGIS was used to spatially map the Chl-a and SM extent around the lake. The LVQ classification model predicted the Chl-a extent with more than 90% accuracy having only one point misclassified out of 14 testing points. The LVQ model prediction for SM resulted in four misclassified points out of 14 testing points with prediction accuracy of 72%. The final spatial zoning maps for Chl-a and SM extent in the entire lake could be used to help water use managers and end users design management strategies, and also demonstrates a relatively low cost WQ prediction mechanism that can be applied in developing countries and elsewhere when detailed in-situ monitoring is not feasible.
ABSTRACT. The measurement scale is one of the most important aspects of remote sensing and hydrological modeling studies. Changes pertinent to scale or resolution of input data are reflected in modeling results. This study discusses different types of scales, resampling and rescaling techniques and several potential scale issues as addressed by past studies. Specifically, scale impacts of digital elevation model (DEM) and land use and land cover (LULC) data impacts on hydrological modeling results are discussed. Higher resolution data can accurately describe watershed characteristics but can be computationally intensive. Therefore, it becomes important to select a data resolution, which aides in model running efficiency by not compromising with model accuracy. The discussions in this paper could provide references for development, application, and improvement of watershed hydrological models, and increase accuracy and efficiency of hydrological model simulations and can aid in making informed decisions by watershed managers.
An important method for estimating Crop Water Stress Index (CWSI) is by measuring surface temperature of the canopy. A remote sensing method was used to estimate CWSI of an almond orchard in Paramount farm, California. An aerial remote measurement using MASTER (MODIS/ASTER) thermal band data used to measure canopy temperature (Tc). The empirical relationship for canopy- air temperatures difference (Tc-Ta) versus Vapor Pressure Deficit (VPD) represents the crop water stress quantitatively. The results implied that the average value of CWSI for well-irrigated (non-stressed) almonds is 0.24 while the almond yield is affected when the average CSWI values for stressed crop due to lack of irrigation is greater than 0.5. The difference in crop canopy to air temperature (Tc-Ta), measured was negatively related to the VPD [R2=0.96 and p<0.0001]. However, the relationship between (Tc-Ta) and VPD used to develop a non-stressed baseline equation for almonds, which estimates CWSI. Determination of CSWI is useful for irrigation scheduling and water management.
Remotely sensed images including LANDSAT, SPOT, NAIP orthoimagery, and LiDAR and relevant processing tools can be used to predict plant stomatal conductance (gs), leaf area index (LAI), and canopy temperature, vegetation density, albedo, and soil moisture using vegetation indices like normalized difference vegetation index (NDVI) or soil adjusted vegetation index (SAVI) developed with near infrared (NIR) and red bands. In this study, we present results of those analyses for two study sites with different plant species: 1) a managed loblolly pine (Pinus taeda L.) forest in coastal North Carolina for canopy temperature and gs and 2) a managed Blueberry (Vaccinium corymbosum) orchard within a natural forest in coastal Georgia (Z-Blu orchard) for the LAI. An Object Based Image Analysis (OBIA) technique was employed on the Z-Blu orchard to distinguish the forest species and establish their correlation with LAI using ground-truthing. Similarly, we used OBIA technique for the forest speciation on Turkey Creek watershed at Francis Marion National Forest site in coastal South Carolina with ground-truthing. Both classified images yielded 80% classification accuracy based on field verifications. Similarly, >90% correlation was obtained for the LAI map developed for Z-Blu orchard site plant speciation. However, for the NC pine site, the correlations were poor, with R2 values of 0.33 and 0.26 for gs v/s Landsat Middle Infrared (MIR) and gs v/s Landsat Thermal Infrared TIR models, respectively. This study on advanced image processing approach for forest speciation and ET parameters prediction/estimation can be a basis for similar other studies in the region.
Geographical Information System (GIS) based groundwater quality mapping has been carried out with the help of hydrochemical data of Jammu district, Jammu and Kashmir State. Groundwater quality for drinking water purposes was analyzed by considering the WHO (2004) and ISI (1991) standard. A total of 100 samples were collected from different geological formations of the study area. The groundwater samples were analyzed for the major chemical constituents viz. Ca2+, Mg2+, Na+, K+ (cations); HCO3-, Cl-, F-, SO42-, NO3-, SiO2 (anions); pH, EC, TDS. In the study it is found that 80 samples in the pre-monsoon and 85 samples in the post-monsoon were under the hard water category (Sawyer and McCarty 1967). The purposes of this investigation were to provide an overview of present groundwater quality to determine spatial distribution of groundwater quality parameters and to map groundwater quality in the study area by using GIS. The ArcGIS 9.2 was used for generation of various thematic maps and ArcGIS spatial analyst to produce the final groundwater quality map. An interpolation technique, ordinary kriging, was used to obtain the spatial distribution of groundwater quality parameters. Five thematic maps with parameters such as TH, TDS, Ca2+, Mg2+ and NO3- having desirable and undesirable classes were integrated using the overlay method and groundwater quality map for drinking purposes has been prepared. A salinity hazard map was also prepared after generating contours which show the regions with low, medium and high salinity hazards. The groundwater quality for drinking purposes was integrated with groundwater quality of irrigation purposes by spatial analysis. Integrating groundwater quality for drinking purposes and irrigation purposes pictorially representation shows desirable zone for drinking, desirable zone for irrigation purposes, desirable zones of both drinking and irrigation purpose and undesirables for drinking or irrigation purposes.
Understanding how the land use change influence the river basin hydrology will enable planners to formulate policies to minimize the undesirable effects of future land use changes. Land cover changes increase impervious ground surfaces, decrease infiltration rate and increase runoff rate, hence causing low base flow during the dry seasons. Efficient tools such as satellite remote sensing and Geographic Information System (GIS) are currently being used to manage the limited water resources. The need for spatial and temporal land-cover change detection at a larger scale makes satellite imagery the most cost effective, efficient and reliable source of data. The ability of GIS makes it an important and efficient tool for spatial hydrologic modeling. In this study, Satellite data and GIS were integrated with a spatial hydrological model to evaluate the impacts of land development in the Upper Bernam River Basin of Malaysia. HEC-1 (Hydrologic Engineering Center) model was calibrated and validated using actual flow data from the outlet of the watershed. The model performance was checked by means of four criteria viz., mean absolute error (MAE), root mean square error (RMSE), Theil’s coefficient (U) and coefficient of determination (R2) obtaining values of 0.14, 0.18, 0.097, and 0.86, respectively. From the hydrographs, it was found that the change in peak flow between the years 1989 and 1993 was 28% while it was 11% between the years 1993 -1995. The reduction of the time to peak was 7% for the same years. The model can be run for any future land development plans to investigate the hydrological impacts in order to avoid the shortage of irrigation water and mitigate the risk of floods occurrence.
Direct monitoring of stream water chemistry is an increasingly important tool for securing stream water quality and assessing stream ecological functioning as it relates to overall ecosystem health. Such monitoring is often discontinuous in spatial extent and, thus, needs to be interpolated at unsampled locations if the desired end product is a continuous map of stream water chemistry. Recently there have been major advances in the use and development of geostatistical methods (such as kriging) for interpolating between observations of stream water chemistry within stream networks. This study investigated the influence of distance definition on interpolation of synoptically collected stream water chemistry samples. In particular, we developed a new methodology for adjusting instream distances between stream water chemistry observations such that instream lakes (which are ubiquitous in northern, boreal landscapes) are explicitly accounted for in geostatistical interpolations. The methodology developed was tested using stream chemistry data for five different constituents coming from synoptic sampling campaigns conducted across four boreal Swedish catchments during two distinct seasons. The ability of this new, lake adjusted instream distance (LAID) to produce interpolated maps of stream water chemistry was compared to that of traditional Euclidean distance (ED) and instream distance (ID). The results indicated that using LAIDs in this boreal landscape tended to improve interpolation compared to the other distance definitions considered. The grade of improvement, however, tended to vary between the constituent, watershed and season considered suggesting that the influence of instream lakes on water chemistry is quite variable in this landscape throughout the year.
A grid-based distributed hydrological model WetSpa, compatible with ArcView GeographicInformation Systems (GIS), was applied to the 7,860 km2 Upper Suriname River basin. Modelparameters were derived from a digital elevation model (DEM), land use and soil type map ofthe basin. These parameters and the observed daily meteorological data (1978-1983) wereused (1) to tests the performance of the WetSpa model to a large tropical basin, (2) tosimulate water balance and outflow hydrographs, (3) to identify the different flow componentsand (4) to study the most sensitive model parameters for the study catchment. The statisticalmodel evaluation results indicated that the model has a relatively high confidence and cangive a fair representation of the flow hydrographs and the water balance for a complex terrain.The use of daily observations instead of hourly observations and the lack of othermeasurements of the hydrological processes (e.g. groundwater flow, infiltration) tocalibrate/validate the model may have caused the large errors in low flows and high flows.The deviations between the observed and simulated flows may also be caused by the lack ofa good representation of the meteorological conditions in the study area. The WetSpa modelalso provided insight into the main flow processes during the year. The most sensitiveparameters for this basin were the interflow scaling factor ki, the groundwater flow recessioncoefficient Kg, the initial soil moisture K_ss and the correction factor for potentialevapotranspiration K_ep.
This study analyzed daily maximum streamflow data of each month from three gauge stations on Cekerek Stream for simulation using stochastic approaches. Initially non-parametric test (Mann-Kendall) was used to identify the trend during study period. The two approaches of stochastic modeling, ARIMA and Thomas-Fiering models, were used to simulate monthly maximum data. The error estimates (RMSE and MAE) of predictions from both approaches were compared to identify the most suitable approach for reliable simulation. The two error estimates calculated for two approaches indicate that ARIMA model appear to be slightly better than Thomas-Fiering. However, both approaches were identified as appropriate method for simulating daily maximum streamflow data of each month from three gauge stations on Cekerek Stream.
Modelling of contaminant transport on watersheds is a problem of concern with regard to the protection of water ecosystems. This paper presents a numerical model, using the finite element method, to simulate contaminant transport hydrodynamics. Spatial and temporal progression of a spilled toxic substance was compiled for the Macks-Creek basin located in United States. It is shown that the developed tool, based on coupled terrain and surface runoff models, provided adequate results in terms of runoff velocities and height. Results from this compute served as an input of a computer based system that we developed to simulate the real-time contaminant behaviour and spreading since its spillage. The developed model can be used in both simulating and forecasting modes. Obtained predictions are of practical relevance to hydrologists and water resources managers for applications in environmental management strategies during infrastructure and urban planning phases as well.
Understanding and managing water resource problems involves complex processes and interactions within the watershed surface and subsurface. The imposition of total maximum daily load (TMDL) regulations on the pollutant influx to a watershed has created a strong demand for new assessment tools. The spatial scales relevant to transport of the pollutants may span many orders of magnitude, ranging from field plots to regional hydrological systems. As the demand for and development of watershed modeling capabilities have evolved, geographic information systems (GIS) in tandem with remote sensing technologies have played an essential role supporting both data collection and analysis. This paper reviews the current and future trends of GIS and remote sensing technologies in watershed modeling. The primary focus of this discussion is on spatial data availability and management, and further opportunities for model development.
An understanding of stream network topology is necessary for a landscape-level perspective of stream hydrology and ecology. We present a method for quantifying stream network topology that overcomes computational constraints of DEM-based analysis over large geographic extents. This method converts vector stream flow paths to raster flow paths to predict spatially-explicit stream properties from a network-constrained upstream cell count (UCC) to flow origins. UCC data enable calculations of stream network structure at designated grain sizes and spatial extents. UCC values were strongly related to empirical measures of upstream basin area (R2 = 0.94) and stream width (R2 = 0.65) within the mid-Atlantic highlands, USA, suggesting that UCC data provide a reasonable surrogate for empirical measures of stream size within the stream network. By reducing raster grids to the flow path, the UCC method reduced file sizes by 99% compared to digital elevation models. The UCC method can improve our understanding of fluvial landscape hydrology and ecology by enabling spatial analysis of stream networks over large geographic extents.
Geomorphologic instantaneous unit hydrograph (GIUH) can be used as a transfer function for modeling the transformation of excess rainfall into surface runoff, in which excess rainfall is an excitation (i.e. production function) to the hydrologic system. These models can be used to predict / forecast the temporal variation of the surface runoff at the outlet of ungauged basin, which is useful in the hydrologic / environmental engineering applications. The present study deals with the geomorphometric investigation and provides an efficient solution approach to derive the GIUH based transfer function and thus geomorphologic unit hydrograph (GUH) for the basin. Since, Gomti river basin is ungauged, therefore, to test the effectiveness of the approach two cases were considered. Firstly, the approach was tested on the catchment for which published UH data was available; and secondly, the approach was applied for the Gomti river basin for the derivation of GUH. To verify the derived GUH of the Gomti basin, a comparison was performed with the synthetic unit hydrograph (SUH) obtained from the Central Water Commission (CWC) procedure. Based on the comparison of the result, it may be revealed that the GUH with dynamic flow velocity of 0.68 m/s was close to the SUH.
Researchers often use only surface drainage (imperical method) as a tradition to locate sites for installation of water quality monitoring wells, and soil sampling profile for pollution studies in urban areas. They neglect that structural development may have altered the natural surface drainage. This paper demonstrates an advanced method; combining surface imperical method with a three-dimensional groundwater flow model and particle track analysis. This was performed using transient MODFLOW and MODPATH codes, as a first step in a pollution study of Owerri urban areas. Result was more reliable, showing diverging flows at Orji and Nekede auto-mechanic villages (MVs) due to relatively high elevation with the surroundings, and southward flow at the Obinze municipal waste dump (MWD). There were vertical flows at the sewage dumps (mangrove swamp areas) between Egbeda and Umuapu, and convergent flows along stream valleys. Particle tracking show an extended capture zone lying southeast. A total of eight water quality monitoring wells (WQ) was then recommended according to the directions of groundwater flow and particle releases from three tracking wells (ABC). WQ1: SE of Orji MV, WQ2: SE of Nekede MV, WQ3: south of Obinze MWD, WQ4: west of Egbeda, and WQ5: east of Umuapu. WQ6: NE of well A, WQ7: SW of well B, and WQ8: NE of well C. Soil sampling profiles; MVs: SE, MWD: South and sewage dumps: East and West. Distance of monitoring wells will vary according to the proximity of human settlements and shallow domestic wells (37m-55m) to the sites.
The Rio San Pedro sub-basin runoff was estimated using the curve number method (NRCS-CN) applied to hydrologic response units (HRU’s), derived from remote sensing and GIS analysis. The sub-basin (around 2900 km2) was delineated from digital elevation models (DEM), that also were used to obtain the slopes in the study area. A landscape characterization (overall accuracy > 80%), based on Landsat ETM+ imagery, was obtained using standard classification methods, and together with a rainfall data series, were the input information for the sub-basin discretization in HRU’s and the runoff calculation. Seventeen HRU’s were obtained, that can be arranged in three main groups. HRU’s associated to forest and high relief areas, representing 2/3 of the total area and contributing up to 71% of total runoff. The land covers related with human activities integrate a second HRU’s group, contributing with 20% of runoff, although they represent less than 15% of the area. Finally wetlands and aquatic surfaces, not contributing to runoff, are the third HRU group. Because of the measures of accuracy correspond to good agreement between the model and the reference data, the HRU’s approach that retains the spatial heterogeneity, is considered a fine approximation for the San Pedro sub-basin runoff assessment that can be integrated to the development of environmental management programs.
Recently, in the year 2006, 2007 and 2008 heavy monsoons rainfall have triggered floods along Malaysia's east coast as well as in different parts of the country. The hardest hit areas are along the east coast of peninsular Malaysia in the states of Kelantan, Terengganu and Pahang. The flood cost nearly millions of dollars of property and many lives. Foods are considered to be one of the weather-related natural disasters. Many methods exist to provide qualitative estimations of the risk level of flood susceptibility mapping within a watershed. This paper presents construction of a flood susceptible map for presumptive flood areas around at Kelantan river basin in Malaysia using a statistical model and GIS. To evaluate the factors related to flood susceptible analysis, a spatial database was constructed from a topographical map, geological map, hydrological map, Global Positioning System (GPS) data, land cover map, digital elevation model (DEM) data, and precipitation data. An attribute database was also constructed from field investigations and historical flood areas reports for the study area. Logistic regression model was applied to determine each factor’s rating, and the ratings were overlaid for flood susceptibility mapping. Results indicate that flood prone areas can be performed at 1:25,000 which is comparable to some conventional flood hazard map scales. The flood prone areas delineated on these maps correspond to areas that would be inundated by significant flooding. Further, risk analysis has been performed using DEM, distance from hazard zone, land cover map and damageable objects at risk. DEM was used to delineate the catchments and served as a mask to extract the highest hazard zones of the landslide area. Qualitatively, the model seems to give reasonable results with accuracy observed was 85%.
The Manning’s roughness coefficient (n) is commonly used to represent surface roughness in lumped and distributed hydrologic models. Model parameter sensitivity studies identify runoff response to be sensitive to Manning’s n changes. For large watersheds, modelers typically use land use / land cover datasets to assign Manning’s n values based on the use or cover class (e.g., residential, impervious). Although this approach is expected to introduce errors to the simulation results, studies have not adequately assessed the occurrence or magnitude because of the challenge of producing an accurate Manning’s n map to compare to a map produced by the land use / land cover approach. This paper presents a watershed scale assessment of the hydrologic model error incurred by use of land use / land cover datasets to estimate Manning’s n. A digital dataset of Manning’s n is generated by manual inspection of aerial photos for a 23 km2 watershed. Manning’s n is also estimated using the land use classes in the National Land Cover Dataset (NLCD). Up to 50% difference in the magnitude and variation in spatial distribution of Manning’s n values is found in more than 90 % of the study area. The differences did not translate into significantly altered runoff responses (hydrograph magnitude: 9 % to 22 % relative peak discharge difference and shape: 2 % to 18 % relative time to peak difference) from 3 storm events at the watershed outlet for a lumped model (SWMM) and a distributed model. However, these differences are significant (up to 75 % relative peak discharge difference and up to 300 % relative time to peak difference) at the subcatchment levels and showed increasing trend in deviation of the hydrograph peaks with increased Manning’s n deviation. The results of this study suggest that the use of NLCD-defined Manning’s n values is acceptable for medium to large watersheds.
