OBJECTIVES

1) Produce detailed vegetation maps of the valley using remote sensing/GIS, depicting the spatial extent and biogeophysical condition of invasive and native vegetation communities in Makaha Valley.

2) Determine the spatial-temporal variation of the different components of the hydrological cycle and their relations with stream flow in Makaha Valley.

3) Calibrate and validate a watershed model to be used as a management tool and in the watershed management decision making process by different stake holders to enable them to predict different short- and long-term land management scenarios and impact of extreme weather conditions.

APPROACH

The proposed work involves four main phases that complement each other.

1) Literature Review: This phase will investigate previous databases for stream flow, weather, aerial photos, watershed research and modeling efforts specific to Makaha or other small island ecosystems similar to Hawaii. This phase has the following steps: i) Historical stream flow, weather, and groundwater pumping. Review previous research work on watershed management; the uses of integrated small watershed numerical modeling applicable to small island conditions; and Analysis of historical aerial photos and satellite images.

2) Physical characterization of Makaha Valley Watershed: Review the quality of the available watershed data.

3) Field Measurement of the hydrological components: during this phase data will be collected to complement available data through state and federal agencies, i.e. USGS. Weather data: (rainfall, temperature, wind speed) will be used to estimate evapotranspiration potential. Soil physical properties: Infiltration rates: Infiltration rates throughout the watershed will be determined using double ring infiltrometer. Hydraulic conductivities: Similar soil samples will be used to determine the saturated hydraulic conductivity of the main soil layers. Continuous monitoring of soil moisture content using capacitance moisture probes during the dry and wet seasons. Bulk density: Bulk density will be determined.

4) Vegetation Characterization with Remote Sensing/GIS Techniques. We will assemble and develop a multi-year, temporal satellite image data set consisting of three images acquired in the same. From the assembled image data set, we propose to derive vegetation classification maps depicting the historical and current spatial extent of invasive and native plant communities and the biophysical parameter maps of leaf area index (LAI). We will investigate the severity and patterns of spread of invasive species using the derived time-series maps and relate them to the components of habitats (e.g., topographic factors and soil physical properties) in order to better understand the infestation mechanisms.

PROGRESS
2005/10 TO 2006/09

The main goal of this work is to use Makaha Valley to develop an integrated decision support system at the watershed continuum by combining multi-scale field experiments with a strong numerical modeling and geospatial component to help understand hydrological processes and their variability under different controlling parameters, i.e., land use extreme weather conditions. This project has hydrological and remote sensing components. We are detailing the progress made in these two sections as follows:

1) Hydrological component of the project

Physical characterization of the Watershed. 6 field meteorological-hydrological monitoring station instrumentations were installed in the watershed. Data have been collected from these sites since August 2005. Data includes rainfall, soil water content, and other meteorological parameters, as well as soil sampling and testing at multiple locations. Soil volumetric water content is being monitored with capacitance sensors at depths of 20 and 80 cm. A laboratory calibration was conducted for the soil moisture sensor to calculate site specific soil water content.

Data analysis. Field data have been downloaded every month. These data have been analyzed, summarized and some of them were used to calculate potential evapotranspiration. Soil moisture data summarized the rainfall data and soil moisture data collected over the same period to qualitatively assess the spatial and temporal variability of rainfall. The results of these analyses were presented in class presentations.

2) Vegetation characterization with remote sensing/GIS techniques. During last year, we focused our activities on two major topics: Classification and validation of the QuickBird and Ikonos images, Theoretical determination of spectral separability among native and nonnative plants. Both the QuickBird and Ikonos images have successfully been classified to their respective limits based solely on spectral signatures. The work resulted in six separable classes, two of which have been identified through intensive field assessments: 1) Ohia-dominated vegetation formation and 2) Kukui vegetation community. Identifications of the other four classes are in progress. Our remote sensing field team located eight field plots and collected the following vegetation characterization data at each site as part of class identification and validation activities: DBH, canopy cover, and species compositions. Spectral separability among six native and nonnative species at leaf scale has been determined successfully. The species studied were: Koa, Ohia, and Lama as native species, and Strawberry guava, Christmas berry, and Coffee. We have also been working to "scale-up" the spectral separability work to a canopy level. A series of intensive field work were conducted to collect biogeophysical parameters of the six plants, including tissue optical properties, soil reflectance, LAI , NPVAI, LAD, and NPVAD. Measuring LAI and NPVAI were found to be very difficult and thus we are continuing refining the measurement protocol for these two parameters.

IMPACT

2005/10 TO 2006/09

1. This project has a positive impact on NREM's watershed hydrology and spatial analysis teaching program of the Natural Resources and Environmental Management Department.

2. The research site has been used as a laboratory for many of our graduate and undergraduate students to demonstrate several watershed hydrology processes.

3. Data from this site have been used to calibrate watershed hydrologic model with water quality prediction capabilities that will allow us to test different management scenarios related to different water-resources issues, including the effects of land-use changes and the implementation of best-management practices.

4. Other models have been also used to study surface and near-surface hydrologic processes in the watershed. The availability of these functional models have been important to the teaching and extension programs at Natural Resources and Environmental Management Department.

5. Specific teaching and research benefits of this project are: i) Advancing knowledge of the regional hydrologic system; ii) advancing understanding of hydrologic processes; and iii) Providing water-resources information that will be used by multiple parties for planning and operational purposes.

PUBLICATIONS (not previously reported): 2005/10 TO 2006/09

Mair, A, A. Fares and A. I. Elkadi. 2006. Evaluation of The Effect of Groundwater Extraction and Long-Term Weather Patterns on The Makaha Valley Streamflow. JAWRA Sustainable Watershed November-December Issue, 2006.