Introduction

Crop growth and yield are determined by a number of factors such as genetic potential of crop cultivar, soil, weather, cultivation practices (date of sowing, amount of irrigation and fertilizer) and biotic stresses. However, generally for a given area, year-to-year yield variability has been mostly modeled

(* Present Address : Indian Institute of Remote Sensing, Dehra Dun 248 001, India; email [email protected])

Satellite Remote Sensing and GIS Applications in Agricultural Meteorology pp. 263-289

through weather as a predictor using either empirical or crop simulation approach. With the launch and continuous availability of multi-spectral (visible, near-infrared) sensors on polar orbiting earth observation satellites (Landsat, SPOT, IRS, etc) remote sensing (RS) data has become an important tool for yield modeling. RS data provide timely, accurate, synoptic and objective estimation of crop growing conditions or crop growth for developing yield models and issuing yield forecasts at a range of spatial scales. RS data have certain advantage over meteorological observations for yield modeling, such as dense observational coverage, direct viewing of the crop and ability to capture effect of non-meteorological factors. Recent developments in GIS technology allow capture, storage and retrieval and visualization and modeling of geographically linked data. An integration of the three technologies, viz., crop simulation models, RS data and GIS can provide an excellent solution to monitoring and modeling of crop at a range of spatial scales.

In this paper an attempt is made to introduce a basic framework and indicate through specific case studies, (a) how RS data are useful in estimating crop parameters like LAI, (b) introduce crop simulation models, (c) how GIS tools are used for crop monitoring with RS data and interfaced with models, and (d) how RS-derived parameters, crop simulation models and GIS are useful for crop productivity modeling. Details on some of the above topics can be obtained from recent reviews, such as Moulin et al. (1998), Dadhwal (1999), Hartkamp et al. (1999), Dadhwal and Ray (2000), Maracchi et al. (2000) and Dadhwal et al. (2003).

LAI estimation using RS-data

The leaf area index (LAI), defined, as the total one-sided leaf area per unit ground area, is one of the most important parameters characterizing a canopy. Because LAI most directly quantifies the plant canopy structure, it is highly related to a variety of canopy processes, such as evapotranspiration, light interception, photosynthesis, respiration and leaf litterfall. RS-based LAI estimation would greatly aid the application of LAI as input to models of photosynthesis, crop growth and yield simulation models, evapotranspiration, estimation of net primary productivity and vegetation/ biosphere functioning models for large areas. A number of techniques for space borne remote sensing data have been developed/tested, ranging from regression models to canopy reflectance model inversions with varying successes, which include (1) statistical models that relate LAI to band radiance (Badhwar et al., 1986) or develop LAI-vegetation index relation (Chen and Cihlar, 1996 and Myneni et al.,

1997), (2) biophysical models like Price (1993), and (3) inversion of canopy reflectance using numerical model or LUT based model (Gao and Lesht, 1997, Qiu et al, 1998, and Knyazighin et al., 1998).

Myneni et al. (1997) developed a simple approach for estimating global LAI from atmospherically corrected NDVI using NOAA-AVHRR data. One-or three-dimensional radiative transfer models were used to derive land cover-specific NDVI-LAI relations of the form

LAI = a x exp (b x NDVI + c) where, coefficients a and c are determined by vegetation type and soil.

Chen et al. (2002) have described relations using NOAA-AVHRR simple NIR/Red ratio (SR). These equations are vegetation type dependent and are being used to generate Canada wide 1 km LAI maps every 10/11 day. The equations are summarized in Table-1 and require a background SR that is season dependent as an additional input. In case of another high repetivity coarse resolution sensor, VEGETATION onboard SPOT satellite, use of SWIR channel is made to compute a new vegetation index, namely Reduced Simple Ratio (RSR). RSR reduces between vegetation and understory/background effects, thus making possible use of simplified equations for retrieval of LAI (Table-1).

Table 1. Equations for obtaining regional LAI products from atmospherically corrected data from NOAA-AVHRR and SPOT-VEGETATION (Chen et al, 2002)

Sensor

Vegetation Type

Model

NOAA-AVHRR

Coniferous forest

LAI = (SR - Bc)/ 1.153

Deciduous forest

LAI = -4.1xln[(16-SR)/(16-Bd)]

Mixed forest

LAI = -4.45 x ln[(14.5 - SR)/(14-Bm)]

Other (crops, scrub etc.)

LAI = -1.6 x ln{14.5 - SR) / 13.5]

SPOTVEGETATION

LAI = RSR/1.242

LAI = -3.86 ln (1-RSR/9.5)

LAI = -2.93 ln (1-RSR/9.3)

LAI = RSR / 1.3

Bc, Bd, Bm are background NDVI for coniferous, deciduous and mixed forests, respectively.

Bc, Bd, Bm are background NDVI for coniferous, deciduous and mixed forests, respectively.

Using MODIS data, onboard TERRA (launched in Dec. 1999), it is now possible to obtain operationally generated eight-day composite 'LAI product', at a spatial resolution of 1km, which incorporates model and look-up-table based LAI retrieval algorithms (Knyazighin et al., 1999) as a part of MODLAND. However, there is a need to validate this product, before it can be utilized in operational applications. Pandya et al. (2003) describe results of a study to develop small area LAI maps using IRS-LISS-III data using field sampling and regression approach and using the generated maps to validate MODIS LAI product. The atmospheric measurements of aerosol optical thickness and water vapour content were performed concurrently with the LAI measurements at the time of satellite acquisition and were used to convert digital numbers into the ground reflectance. These images were geo-referenced and the fields within the region of interest, where LAI measurements carried out were identified on images. Using NDVI of these fields, empirical models based on site-specific NDVI-LAI relation were developed (Figure 1) and used to generate LAI maps for each acquisition and study site. The LAI images were aggregated to 1km spatial resolution and compared with MODIS LAI product and results indicated significant positive correlation between LAI derived from LISS-III data and MODIS data albeit with a positive bias, in the MODIS product (Figure 2).

0 0

Post a comment